 Shall we kick off ladies? If we can make our way to our chairs that would be good good morning all Anthony you've got it. I've got people trying to work out your acronyms So I'll get you to tell them at some point Okay Yeah Monday Monday okay Raj apologies for the delay starting but again just a reminder of whose country you're on Warundry country and also the values that we talked about yesterday around caring and sharing and respect and an acknowledgement of your own unique and special qualities and you all were thinking about this last night your desire to come together Make sense of this process Get some clarity around all the moving parts and identify the synergies and just commit to working together If you think that you've tricked me over there moving around I can see where you are And my deadly sister here Was there That's all right, but you know what I notice of you Same guys a different top No, don't worry about it. I'm the same. I'm the same now There'll be few Conversations that you can have throughout the day with yourselves, but at towards the end of the day I want us to start thinking table by table About what you think is needed to improve a process If that makes sense, yeah, how do we make this? Real for everybody there were some questions the issues that came up yesterday in terms of we can't assume All Aboriginal people struggle to understand some of the the terminology Some Aboriginal groups will just pick it up and rumble that they'll understand it completely They're capable of doing what they want to do, but there'll be needs to adjust in certain environments, but so I was saying that one one What works in one community won't necessarily work in another community So we just need to be mindful of that and the only thing I'll say is that at some point all of this language and acronyms and Words words words, they've got to be put into some sort of one-page picture So that everybody can see where it all fits pretty much and that would be helpful for a lot of Aboriginal people, I'm sure Okay, now So about that or not so about that or not. Good morning Raj Yeah Be polite and courteous and there's many languages as we can You're going to present today on The NRLs Upcoming season And you'll probably report a little bit on the Big Bash final for Saturday night. I wish the Brisbane heat got into No Raj is going to talk to us today about Just another acronym Iris services and sensitive data environment. So I'll get you to start coming up Raj Raji Raj Rajeev Yeah Rajeev Raj is your homie name. Yeah Rajeev Samaraj is a senior research fellow and the program coordinator for data and analytics at the Melbourne Institute applied economic and social research leading institute for applied economic and social research within the faculty of business and economics Rajeev has over 13 years experience in data analytics data visualization and developing statistical methodologies for data analysis past experiences involved transition of engineering concepts through the health Science as seen through the work. He did with for his phd In deriving meaningful quantifiable information from image data presently Rajeev Is translating his engineering experience into a to innovative data analytics and data visualization and the social sciences including oversight of Melbourne Institute data lab Lead for the foundation fellow training program and his integral and his integral to the iris Integrated research infrastructure and social service platforms. So we'll hand you the floor Hi everybody Good morning. I have the pleasure of Where's the little do the key? I have the pleasure of being your first talk today. So I'll try not to make it I'll try to keep it as upbeat as possible to get the pace going for the rest of the day Before I begin, I'd like to acknowledge the country We're on today, which is the Wurundjeri people But I'd also like to acknowledge the the Boon Wurrung and Boon Wurrung people of which the land I live in Because we all work from home for the past three years and a lot of what I'm going to show to you today was inspired Was done in this land and the land I live in as we work from home um So as uh Grant mentioned, I'm I am an engineer Um, so that's a disclaimer. I am not a social scientist and I'm not In the ha s s and I am not unfortunately. So this was a Joke we used to have there right now. Whenever I did my phd at monash, uh, this was Stuck on the door outside where all the phd students were working and we all thought it was funny, but um So I don't understand But I want to start by saying, um, so I guess yet another acronym. Um, what is iris? But no, I don't need to have a Michael and Wojtek did Too excellent talks yesterday talking about the Integrated research integrated research infrastructure for social sciences. Um yesterday um, so I'm not gonna start start on that but um We the iris project essentially splits into six work packages Of which is yet many more acronyms But around, um, you know, a lot of work packages around, you know, survey design Uh, the geo social elements that we heard about yesterday Um, also the curation aspect as well, but today I'm My work package is actually a part of the work package. Actually Is around sensitive data analysis demonstrator And what this work package is about is about so we have all its integration work we're doing And My work package our work package is about okay now How do we take that work that we're doing and how do we move it to a sensitive environment if we're using sensitive data A lot of time we cannot use sensitive data in our laptops on our own computers So we're going to take it to an external environment. So a lot of that work is done around how do we demonstrate that And for phase one we had when did some work around Looking at some of the requirements people had around working with data Some of the pitfalls when it comes to accessing data, especially sensitive data, um, and you know, some of the issues there Um, and now we're currently working towards the development phase, which is around, you know Architecting what that integration exercise looks like in a sensitive environment And then moving on to how do we create a pilot of that The message phase one will also touch on you know future work at the end of the end of the talk So I thought I'll start here with what I mean about sensitive data And no better place to start than Starting with a privacy act um, and so, you know personal information, um really means information about Identifiable information, um, and so what this really means, you know individuals name um signature address Record employee record information a few other things Um, so that's personal information, but there's also sensitive information And what this is defined is to be around your racial origin political opinions philosophical beliefs And a range of other sensitive information. So Now this is very sensitive data And where social scientists One access to this data is around making Perform creating undertaking research analysis and informing social policy and evaluating social policy. So As as you can understand, there's obviously a lot of a lot of issues with accessing this type of data Especially and we've talked a lot yesterday about, you know, the data owners were the rightful owners of that data We are Um, you know everybody everybody owns their own data But it's about how can we better use that data still enable research, but also be like be be um What do you call we we all know the owners of the data so Our experience usually with data is so we have a data custodian. They collect the data And They usually link the data to another another data set. Um Then they undertake some work around Curating the data to create something that's a bit more research ready And once that's done, that's what you call that's what you get the data set And this data set Usually is not something we can access right so depending on what information goes and especially with sensitive data So usually what happens is there's some treatment is done to this data Before we get access to it And so you so what we call open data we have we get access to a data we can just download it from the website But there's usually some licensing that we have to agree to We have to attribute where we get the data from and that's where we get that's how the user gets access direct access to the data Another model is the limited access model. So In this model you have specific agreements that you have to sign that you say you will you know use this data in a specific way um And usually have to create a project summary, you know, what what what do you intend to do the data for what are you going to do? do you have its experience and Usually there'll be a reduction of the level of detail in the data. So addresses won't be You know revealed They'll all be coded down to a specific geographical level You know ages might be done in brackets And or you might be even aggregate level at a national level or a state level data and you can get access to this data and you can use it in an environment The other method is accessing it through a secure access environment And this is once again, you have more controls again You have access agreements of confidentiality need polls You'll have to undertake some sort of security training Before you can access the data that tells you how you can use the data what you can do with it Also about what you need to do to the data before you can take it out of the environment, which is very very critical Once you get access to the environment Then you have your the data is going to get the data then you can clean your data some more You can then do your analysis and then once again I put treatment at the end there because you'll need to do specific things to make sure you can take only specific counts out of the environment Before you can even take the data out. So for example, this specific rules around You know Each table must have at least 10 counts before you can take it out because you might be able to identify someone within the data Um Other things like you know looking for dominance, which is where if you have you know Data set and you know, you know one value is much much larger than the other Then you'll be able to infer who that person or that individual or the business was to the data you're taking out So you'll be very careful. So once again to do more treatment Before you can take the data out so What is a sensitive access environment? There's lots of words for it. There's lots of names for it. Um Trusted research environment secure data environment data safe haven Um secure data laboratory. I love that one But there's lots of it and they all kind of do Mean um very similar things um Back then uh a long time ago, I guess some of you might remember. I think we still have this actually we have Secure access was you know, you'd apply for a data set and you'll have to go to a specific location They'd do a retina scan fingerprint Give the name, you know, your kids, but certificate And then you get in and then you would access your data set and then somebody will be watching over you You can almost hear them breathing down your neck while you're accessing you do certain things a lot and then you know You can get out and Be like, oh my goodness But that was that was just very physical Secure environments and then we've kind of from then we moved to Trusted execution environments where you'd come up with some code. You'd work with some dummy data You'll give it to someone you'll give your code to someone to your program To someone and then run it on the data for you. You don't actually see anything and then they'll give you the output a bit like table builder really Which we still use And then we from then on we've moved to okay, we can access an environment Looks, you know, looks like a virtual desktop. It even looks like windows has your data in there. So that's more of a modern concept of a data lab More data environment um, I'll be talking a lot more about sensitive data and sensitive environments next week At the summer workshop. So if you're ever if you are in sydney, um, I would like to say Hi to you and you know speak further on this matter, but just wanted to put an ad out there So I guess where I come into the picture Where we come into the picture with iris is we built the melbourne institute data lab middle Sorry grant yet another acronym um But we like the name middle I'll share a funny story. Um So we have middle which stands for melbourne institute data lab, which was the secure environment we built But I sit in the administration team and guess what we call them the meeting We call the middle management team meeting So Yeah, so um, so the middle uh is a is a purpose field environment that we built for analysis secure um social science data specifically to um Undertake analysis using survey data or government administrative data As well as random control randomized control trial data Um, we built it in late 2021 as an established as an agreement between billet issued and a company called second soul, which is a security professional services team in in camber Um, and we also got a lot of support from the university of melbourne. Um in building it Business services legal and risk and whatnot and was also built with funding from the paul ramsey foundation and the new melbourne Um And so what it is Is it's a remote access environment? Uh You the user would go to a web browser. He type of uh a browser. I would call um a web url Um This is how you access it. Um, not gain access well talked about a little later Um, but how you would normally access your data is you go to web browser You'd log in you'd get access to a windows environment It looks exactly like your windows environment, but it's a lot more closed down And then you'd have your um data in a in a drive And then you'd have all your statistical packages You'd need to analyze your data and you want to take all the analysis you want inside that windows environment I was going to put a picture up of it, but I would have just put a picture of What windows looks like and that would have been quite boring. So instead I made a pretty picture of it But the the main thing here is when you need to take data in you need someone needs to approve it Someone needs authorize and bid it bid it and approve it Um Before you can take it out of the environment And you don't have access to things like internet and email and things like that but you know So how do you get access to it? Um typically, um A researcher or a group of researchers would submit a project application And in this application, you'd write what you want to reach what you want the data for what data sets you want or what data sets you're bringing into the environment What are you using it for? What are your questions? What do you what do you what do you what is the research objectives and your plan? And this uh application would get better internally by the team But then it would also depending on what data sharing agreements we have we might have to Engage the data custodians to make sure Can I give access to that data? Can I give such and such person x access data if they say no, then we can't give you access but We haven't had that happen yet um And then you undergo training So we have a security awareness training which talks about you know what you can and cannot do in the data into the environment Um this includes things like you know try to identify yourself in the data set you see you can't do that identify your friends You know, or if you think you can identify somebody then you need to tell us about it So training around you know what you can and do cannot do in the environment We also give the option for data custodians to give us specific training material If they want users to go to additional material as well in on top of what we do because they might have specific things around you Which laws they need to abide by by accessing our data And once a project then we get the project approval. Once that's done we get the user agreements and confidentiality deed pools Signed See if it works. Hi works And then there's two ways you can access the data once through what we call a standalone project Which is where you have your project you have your data And that's it and you have your tools and you know people you that were on your form. They can access the environment And That's it. Nobody else can access the other method is what we call a shared data environment, which is where It's essentially a group of projects that can work together But under a specific research team, so they're essentially done by teams So it might be on we're investigating. We know we're we're a reshared data environment around employment around you're looking at Employment so we're bringing a whole range of data sets on that And the idea is this permitted data sets that you can have and different projects can access different data sets And once you have that you have the output clearance process Which means in all projects much all outputs must be Fitted by the data custodian and or someone delegated before they can be cleared out of the environment So let me talk about a bit more about the shared data environment because this is one of the things We really like about the environment is we want to build a community of practice and this is done basically through an internal wiki part of it's done through a wiki where We create a range of The users create material on the wiki and they can share within their shared environment without the projects. They can share code They can share Things they learned about a data set, you know, I was analyzing a data set and I found our income brackets were really useless or I found something was wrong and I wrote some code that Basically found a way to treat those income brackets to the way I want to do something Or I found that I was analyzing census data and I didn't like the way Depending on how the question was posed in the form that means there will be some issues with the data We're actually collecting about The household So those sort of issues I can write down in a wiki and then I can share with my colleagues within the environment And anybody uses that data set There's a data so every each data set has their own wiki And depending what you cannot what which data you can and cannot see you get access to different wikis So but then you can learn more about the data And you can access sample code and that you can then use that on your environment on your own project Another thing I should mention is shared data environments also mean you can have your own governance framework around who gets access to that project For obvious reasons if your project is about a specific Field of research employment poverty, then there'll be specific things around Well, if you're not doing, you know, there'll be mechanisms around, you know, who can access that data and that project and that those resources Another thing within the shared data environment is this concept of a research ready data set So we as well engineers And social scientists also we always find that we always do We get access to data and we sometimes we find ourselves doing the exact same thing on a data set before we need to analyze it We all do it, you know, we always take it. We always transform it a specific way And we're all doing it and there's all this lost Time in effort in creating that So what we want to do is if we find that we're always doing a specific thing, then we can go halfway And we can document what we do and then we can create a program In multiple languages, you know, state or python That can then mimic you to get to that point in time So you get the as provided data, which is what we call the data we get from the data custodian And then there's a step that takes you to the research ready data And so with the research ready data, some of the key features we want to have is, you know, it to be available for broad research Not just for a specific thing, right? Because then that the researchers themselves can do that bit But we know that we we make sure the data is meaningful And it's usable for research and statistical purposes You know, it should be accessible and be a lot to be reused Um, and there's sufficient robustness checks that we've done and we've documented and how we've done that data quality checks Data quality statements that we would incorporate if the data custodian hasn't provided them to us Um, and any measures of relevance specific to that shared data environments if you're investigating more poverty Then there will be poverty measures included into it. I mean new variables included into it um And then almost other relevant programs, um or code um, so I guess, um one of our biggest the the shared data or the side at all is what's called It's called the breaking down barriers shared data environments. This was the one that was funded by the poor emcee foundation And this was the first one and the idea behind this shared data environment is to inform And shape policy around understanding disadvantage and poverty in Australia. So we've done quite a few reports. Um Within the breaking down barriers project And the shared data environment we're working towards currently with a range of other data custodians ABS, ATO DSS or services Australia depending what you talk to um, you know, bringing a range of admin data into the environment We're also talking to a few other service providers like brotherhoods and Lawrence brotherhoods and Lawrence and um, you know, bringing out the data sets in as well um And as part of this, you know, we're building data sets where we're bringing data sets Um, we're undertaking reviews on, you know, specific topics relating to poverty and we're also, um Writing longer pieces of work called we call demons demonstration reports Here's an example of what it looks like what such a wiki would look like. Um, this is uh, you know the wiki Oh, this is never pointed what you can see on the left hand side of the of the wiki there No, there's specific, you know, we can get specific details about the data set um You know the how you get access to it, but we also give what we call the data snapshot Which tells you, you know, what's the name of the data? What's the overview of what the data set is who do you contact if you want to get access to the data? Very important Um, you know, what's the sampling framework? How is the data collected? um, you have the coverage, um, and you know any other data quality So it's like a very brief snapshot And then we do like a deep dive into the data. We know what variables are there So this is the data set we created for this project using census data using cross sectional Individual level census data. So Then, you know, we have what population measures are captured, you know, what question was used what was census question was variable How was it developed? So we capture all that information and put it into a big into a big wiki form um And so with this, um, we have a data governance model, um, that we are currently working on at the moment around, you know There's a data governance body which sits around, you know, between the data custodians and the data Uses so that sits around, you know, how is that access? Um, how do we look at that access, you know, what how do we talk to the data custodians around, you know Who gets access to data because not everybody can get access to the data. Let's be honest, right? So you have to You know, you know the data uses or the potential data users You really need to state why they need the added data what they need and what, you know What are they going to do with it and for how long? So that governance body is around making sure that we have that conversation and we have that vetting process to To vet that We also sit with a couple of the other middle governance structures, you know between, you know, our technical advisory group and executive council and information security council. That's related to the security of the system Oh, I forgot to mention the security bit. So yeah, so we um The system is uh We got our IRAP certification last year for protected level under the thrilling government standard And we're currently working towards accreditation of the new data availability and transparency act to data scheme So we're working towards accreditation for that at the moment Which is quite a process And so because of the security of the level of data our sensitivity of the data we host we had to get sign off from universities Vice chancellor and the CTO so because we have we have to be represented on Because of the risks associated with the data we host and the platform we provide we have to Be represented on the information security council. So I should have covered on security earlier Um, and under that we have the data stewards. So if you're bringing data, like I said with the breaking down barriers different different data sets um They get to work with what we call data specialists and data curators In creating this research ready data um, so they know what specific focus they need to have And then the data was under that we have the output fitting policies and other policies in there So for iris um what we're looking at for phase one is um, essentially taking the geosocial work that michael and boy tech went through yesterday and taking it out of the Computer environment, which is you know your personal laptop or your workstation at work And putting into a into a sensitive environment and with that comes specific complexities around Internet access, you know, if you want to access an api or you can't so we have to write in specific rules into the firewall specific to your project might be at specific user level to Um around how you can get access to specific api specific um other tools outside outside the environment So for us, it's around working with the geosocial so bringing the scripts in um And then so when a user, you know Goes in and write what they want. They want to the integration work We need to have specific scripts that talk to orange technology orange infrastructure UQ infrastructure and then bring that all back in To our environment so that we can still keep that integration exercise alive um And then we're putting the curation part in through the wiki that I was just talking about um back into the into the project um And then that creates an integrated data set and then the copies of the executed scripts and whatnot And then if you want to make it out you still have to go through that same process again as before That's what we're working towards phase one and what we're looking for the future Is really around okay, so we have an environment. So what? Right, that's what I always keep thinking. There's lots of other environments out there so For us it's about making sure we can get researchers access Through the data they want but also access, you know, specific features on apis and other cloud-based services But also kind of going back to what we call like a data market. So we have cadre which you know, um At a new we also have data place which is by on dc So there's all these data markets where you can access data or you can request data So we need to be able to talk to each other Talk to the data market but through that talk to other sensitive environments as well because right now I don't know like Right now I find it really difficult when I have to go to abs data lab I do my work there And then I get my results, you know finish my work and then I go take it out And then they vet then they when it goes through the vetting process I've already taken up half my counts And then I put it back into my lab and then it's like oh, that's kind of pointless Or if I put somewhere else like that's kind of pointless. So what if there needs to be a way where Sensitive environments need to talk to each other And the key part that that is the credentials or accreditation So if you have ira if you have iso ism five safes, you know, we need to have work towards a mechanism where You need to know, okay. I'm sending something in the middle the data cluster and it's no, okay Yep, they're better. They they have some, you know credentials. So it's safe to send the data. So I feel like We as scientists Talk to each other venues like this and I know the cyber security specialists talk to each other in their symposium But where are the scientists talking to the cyber security folks, right? And working towards, you know, how do we get how do we get to this right? So that's just my last comment there So there's the growing need I want to close with For granting access and enabling better use and reuse of data And there is no one size fits all when it comes to data labs So it's about making access to, you know, it's all based on what project and what work and what data you want And Lastly, you know enabling faster and richer analysis Using a broad range of data sets is critical With minimal risks to the to the data owner. It's critical for our research. Thank you Any questions Dark room in the middle of a of a building to access sensitive data is not that far back. I I did it in 2019 with You still have those rooms and yeah We were evaluating the Melbourne supervised injecting room and accessing Medicare data and pbs and all sorts of very very sensitive data So I wanted to ask that is is middle an accredited data service provider for for data sets curated by Groups like the Australian Institute of Health and Welfare for Medicare pbs and all of this That's one part of the question. Another part of the question is um I presume it operates on a cost recovery model of sorts. So there is a Is cost recovery and in terms of the ADSP. Yes, that's the we're putting in an application. Well once I can get Deputy vice-chancellor to sit down and have a chat with me and then I can Submit it, but we're very close to submitting that and the idea is it we're not thinking as it said as a university of Melbourne solution It's a national application. So yeah, we were open to working with others as well But we're not going in for a data integration Service. No, we're going to let a hw and no easy deal do that. Yeah, and the last question is Who can access middle to work? Is it Internal for University of Melbourne? Is it open to external partners anybody? Yeah, so if you're if you're from a research institute, or if you're in Australia, you can access it That's good. I should have mentioned that. Yeah, thanks Isabel Are you guys thinking of becoming one of those accredited Data facilities under the data availability and transparency act? Yes. Yes. So that's why the ADSP Accreditation, so we're very close to that Finishing that actually finished it. We're just waiting for approval to submit. So, um, yes Thanks. Thanks, Rashi. Um, this is operating now, right? Yes. Yeah, so that I was just wondering about, um the length of the process between when a Researcher conceives of an idea and then when they actually get access and they can start their analysis. What's the What's is it? Do you have a typical time frame or kind of what's the best case scenario between? I'm just thinking about the the Productivity Commission had a report about the Data and digital dividend and this is kind of the counter to this that this, you know, this process Is slow. That's my my kind of get can you give us some kind of real world data about How much time elapses between? The researcher creating their concept and then actually putting it into practice So if we already have the data within the environment Um, we've done it in three four days getting everything running So it once again, it's once you get you once we get the application is about how quickly we get them enrolled under the training And once the training is done, it's literally half a day to get the project up and running. So we've done it within three four days That said we it once again depends on that It's the data is there if we have the data within the environment So that's the main blocker for us. If you need to go and talk to somebody else, we're getting the data that's different Um, so one of the just to say one of our critical user requirements that we had was we have to do it within a week Right, or we have to try our best to do it within a week Because I know sometimes we do that and we would wait a month to get the data So it's it's it goes back to once again to the data if the data is available within the environment or not Hi, I have a question from online from margie smith. Uh, who asked with the data data and transparency act accreditation will you need to have accredited users Why do you for that? Yes, um, so the ondc has two parts to the accreditation. Um System so there's accredited users and then the accredited data service providers. You have to be an accredited user To access the data or to access the data services within the data service provider And um, I should have actually had a sign up there. Um with the data services provider There's three data services that they can provide. So that is secure access They can provide complex, um data integration or the identification or three Um, so we're going for the identification and For secure access, um, but yes, but you'll need to be an accredited user. Yes So we should just make a clarification on that. There's a very interesting part of the data act Which is there, which is none of the people in this room will be an accredited user It's actually your institution Okay, so the user in that system Is the you know the university melbourne's very national university and so forth Um, so there is a and the iodc is coordinating some of this work. Um, the a process of accreditation The institutions have to go through which is actually fairly high High intensity what has to be done in fact That that is sort of underpinning a lot of this in the background here um And that is part of the driver for court, you know, particularly why look at some of those sensitive data environments is This this is kind of an emerging field of how this will actually operate And we're doing some other work in the background to Sort of align some of those processes as well So very nice to get that respectful sharing, you know model that we got to yesterday And changing some of that language, but there is a lot of work going on on You know in different parts of the the overall community that's that's kind of driving this Thanks to you I have another question from online. Oh, sorry It's from julia who asks how long does it take if A i h w has to integrate their data first in your experience of such requests Does that does this include access to abs link data under my nap? Oh m m a d i p sorry under my dip. Yeah, so um with abs. We have to be very careful. So specific data sets we can't access because they fall under the census and statistics act And abs cannot give that data even if they wanted to um So we're working with abs on what data sets we can bring in in terms of the time it's How long is a piece of string? Unfortunately when it comes to data integration because it matters on what day what measures you want What level of data you want a little detail you want unfortunately um We had projects which took six months to get the data integrated first Um, and i'm still waiting with the ato. We're still waiting for data. It's been two years now But mind you that's not just access to data. That's that's separate to bring the data into hosted for everybody else. It's a bit different Yeah, I just wanted to add to steve's comment So so the university accreditation process is is happening now um, I know certainly in the case of the group of eight universities that they are working together um around the accreditation and sharing um information, um about how they're going Some of those institutions are at more mature stages in in preparing their accreditation as users Than others but that information is being shared outside the goa. Um, I'm not sure what's happening um I would say if if if you're a researcher in a university who wants to be able to take advantage of middle One of the things to try to do is to make sure that your university is in the process of seeking accreditation as a data user And for researchers to actually I think lobby Within their institutions for that to be happening if it's not The other thing I think is really important Is for people who have experience working with data like this in environments like this that researchers actually feed into the accreditation process Because there is a risk in fact that accreditation will sit Within it departments or it sort of central service divisions of universities and cyber security divisions And and it's important in fact that that it's also informed by the way in which research practice with these data and in these environments actually happens Just to add to that. Thanks. Thanks a lot mark. Thanks Steve as well. I'm just to add to that, you know, so If the accreditation applies if you're coming through data place if you're coming through that platform if you want access Data made available through that act and through that scheme Um, and there are ways you can access data without that environment. So as well. So just that's not the only way One two, and then we'll have to keep moving unfortunately Just on just on the point of uh accreditation. Indeed the go8s are talking to each other. Um ARDC is also facilitating a national community Which all the universities are invited to and I think at present there's 20 or more universities actually coming to these meetings So, um, there's there's both happening and we're also talking to the go8s and they're also in that meeting So if you want to know more about the ARDC community, please please get in touch That was exactly what I was going to say. So, uh, yeah, please write to us on contact at ARDC Do you do you and join our support group for those trying to become data place data user accredited accredited I wanted that part out because you know, you could mistake us as father and son Thanks. Oh, you okay? Yeah Thank you. Is that yours? Otherwise, I'll do your presentation again now. I'm mindful that we need to catch up on time here. So these um We've had a bit of a shuffle We're going to talk to uh, ryan this ryan perry is going to talk first um, so we switched to presentations around there ryan ryan's uh Uh, let's say activists, but no, you're an archivist. Um at ANU's Australian data archive working with management of longitudinal data data sets in the in the in the archive and developing improved achieving process archiving processes including for handling deposits privacy and risk assessment and archival archival systems social psychologists with 10 years research experience cross psychology management public health including the ANU center of research our social research and methods academic fellow at the university of melvin apartment of management and marketing and mckenzie research fellow at the university of melvin school um The population health ryan is presently working with hasson i rdc team on the iris Yeah, thank you. Um, and first I'd like to acknowledge the Uh, Wurundjeri people the traditional owners of the lands on which we meet today Uh, and where I live and work myself Uh, and kiora koto tenakue. Thank you. Um, so i'm gonna um speak uh, a little bit to um the the processes that we use Um, so i'm going to um speak uh, a little bit to um the the processes um that we use Internally at the australian data archive for handling sensitive data And and you'll notice there's a considerable bit of overlap here, um with what's happening in the the broader iris project Um, so uh, I'm also I'm also going to detail um some of the recent uh, scripts and apps that we've been working on Uh, working towards a self-service um Self-service process for um data owners that are interested in submitting their data to australian data archive Um, so uh, a little bit of overlap with um, rajah's slides. Uh, gladly. I'm it's good to be on the same page um, really all I'll add to that um is to say uh, a critical part of the the the privacy act for us is this idea that um sensitive information is uh, sensitive data is Defined as as that with both um identifying information and sensitive information And for much of the work that we're doing to uh reduce identification reduce some Risks around sensitive data um is targeted at the the uh identifiable side of this so it's it's more straightforward um for us to De-identify the data Which um, you know Sometimes means the sensitive data sensitive information is able to be published but um as was was clearly established yesterday That's that's not always the case. So there's definitely some exceptions That require more manual input from data owners and and from the archivists Okay, so um effectively We are trying to balance access and and utility in the data so not um not Changing it or or or removing too much information. Um, but this of course has to be balanced against Privacy risks due to sensitivity and and identifiability Um, we can modify the data in various ways um, but um, even so, uh, most Sensitive data is published with conditional access um, and the um, I'll I'll explain a little bit more about, um How we do that I think, um So, um Much of the access is is controlled, um manually at the data archive So we have a team who process each request for the data and And match up the the credentials of the person applying for the data With the customized access rules Determined with the data owners input And our system also ensures that there's detailed metadata attached to each of the Each of the data deposits with us meaning that, um That even if the information is is highly restricted or has a lot of access conditions It's still very findable on our catalog system So people can find it understand what's in there and then proceed with with the access process um, all right, so so just in summary we have Two main strategies for managing sensitive data of the Archive, so there's the licensing and access conditions Which I was just talking a little bit about so as well as the access conditions. We also develop a customized data sharing agreement or or license with the data owners And we as archivists we have a range of both automatic and manual risk assessment processes And I'm so I'm mostly going to be Going through a few of these data risk assessment tools So we've been developing a suite of R-based scripts and apps And these are working towards Not only to make the process a little more a little easier Automated for the archivist, but we want this is also to become more of an educational tool For the research community and a means of self-service data risk assessment and de-identification Okay, and as I mentioned So we have access conditions and these also include Publishing different versions of the data so we might publish both a general access with low or Low access conditions and a restrictive version which requires More criteria to gain access to the data And we're also gradually incorporating the the five states which is a more formal framework for restricting access and ensuring that those applying for access are deemed safe people and their Projects are safe and the settings in which they're using the data are safe Okay, so on to our Our scripts for assessing risk and de-identifying data so firstly Our first script that we work with is is really just using a keyword search or key term search of the data labels to the variable labels and names to You know to identify common identifiers so anything to do with names addresses Participant IDs So this is a good sort of automatic Check so that we don't miss anything but as you can probably imagine it throws up a lot of false positives as well Um And we're also trying to Trying to identify Indirect identifiers. So these might include geospatial terms So if there's geospatial information like postcodes, it's going to make the data riskier The ability to identify individuals increases And I'll talk a little bit later about the the importance of of testing for unique combinations of responses so being able to identify someone by combining A range of different demographics that that might be unique in that data Um as part of this we're also assessing the data quality And I'll just mention briefly that That we're automating it as much as possible But the process still includes a lot of manual assessment from the archivists including To evaluate sensitive information in the data qualitative responses And and other quality issues like Consistency in the naming and labeling Okay, so this is just a screenshot of the output from that that keyword Script that I was just describing So something really nice about using the markdown program is that the output of the script is a word document That we can edit you know As we evaluate whether these are true positives or false positives for privacy risks And it's also a Report that we can give to the data owners so that they can approve any suggestions so You can see there So this is a This is a very risky data set And you can see the the keywords that it's throwing up are things like name And I name an id mostly there and then you can see the the archivists Comments and the owner's responses At the other side Um, okay, so that is that's how we deal mostly with them direct identifiers like names and id's It gets a little more complicated when we are assessing The the combinations or the indirect identifiers that I mentioned before So we have again in r using the shiny package We've developed a web app Oh, it's in development Where you can load in SPSS data and then choose various variables from that to To examine How many people have these these different combinations of demographics? So in this example, you can see that there's a lot of relatively unique combinations, so we're interested in Um in combinations where there's three or fewer respondents And you can kind of imagine from this data that if it was a if it was a single workplace or institution Uh, you probably could identify You know the 18 and 19 year olds that were in certain roles in your organization There might be very few or only one of those um, so if there was also Sensitive information in this data set Then you would be in trouble So, uh, this is also, uh, as I said meant Eventually to be something of an educational tool So we're trying to make it Visually quite appealing with the you know the red highlights of when there's potential issues And then there's a second Dialogue is part of the app here Showing you the the frequency of each of the categories so Here you would be able to identify that that some of your More low frequency Occupation categories in their So sales workers machine operators And the other category and the app allows you to Select These categories that you you want to combine together so you might combine them all Into an other category So that they Wouldn't be identifiable anymore And once you've uh, once you've made that combination Does it to the data in the background and then you can go back to your Evaluation screen and you can see that we've removed Almost all of the unique combinations there And in the background there I've combined the h into Categories as well And so from there, maybe you would want to Combine your age groups to even broader categories to remove The remaining unique combinations Okay, and and so one final step so we are expanding that work now to deal with risk associated with panel data And this is also relevant to Data linkage as well So You know the the more pieces of data that you have Across different data sets on the same person the the higher the risk of being able to identify them That you so you might not notice in one wave of the data that there's a An identifiable person in that because you know in wave one they gave more information or in another wave They might have given sensitive information And you haven't sufficiently de identified them in the way of you working with so What we do in these situations is Is first of all limit the number of demographic variables And their categories So we Working with the life in Australia panel We only we only publish the data with with four key demographics and and recategorize Those if there's lots of categories And this is really just to reduce the demands of the The suppression that we're doing on the data So if you have lots of variables and lots of categories Then you have to suppress lots of responses to make sure there's no unique ones anymore And so again where the whether the our script is identifying A combinations with fewer than three respondents or three or fewer And then it's it's suppressing those unique combinations by replacing one of the responses in those unique combinations So if you're the only person with a with a specific Education age gender and Job Then the algorithm will decide to Replace one of your responses with missing data so that you no longer have a unique combination of responses And I'll I'll try and explain as well how the how the algorithm is is determining Which response to suppress there? So this this took us a long time to get it into a you know a coherent paragraph so So we suppress a given demographic value If the combination of the other three Is more frequent than other combinations containing the given demographic so So you might choose the person's education response so If their Age gender and job A really common combination in the data because that means their education is the most Risky or the most likely to identify them because it's As common when you combine it with everything else All right, so just to To leave you with the limitations of what we're working on here so Of course focusing on the Identifiable aspects of the data Is not going to sufficiently Protect all sensitive data So as as was established Yesterday the context And the cohort that you're working with Is important because sometimes the even if it's not Directly identifiable the sensitive data still might might not be appropriate for publishing So this requires a lot of Manual handling from the archivist and working closely with the data owners to establish like what's appropriate um We With the apps struggle to adapt to different populations and sampling Procedures might serve larger datasets are Less risky or if it's context specific like my Single organizational institution example before the risk increases Uh, likewise, um, we don't uh, what the this side of the The risk assessment and de-identification Doesn't approach the research ethics and consent side So that's also an avenue that we are trying to explore more is to ensure that Data owners the outset of their projects Um You know collecting consent appropriately Thank you Any quick questions if you do, uh Hello, oh, sorry To sorry. I said quick The um, do you see it problematic with indigenous people, you know, um When you know Any I just sort of saying do you see it problematic because sometimes say if we go, oh How many indigenous professors there are in Australia? We would know them all basically You know something I just Any I just put that to you just say and say in the community Everyone know that maybe uncle or auntie is Is the knowledge holder there and to deoid and they have sensitive stuff Um, anyhow Yeah, uh, yeah, it's it's a very good very good point. Um, if you have, you know, very very few numbers of certain demographics Um in your sample and your population then Uh, you know, I think one of the main dangers from this is that you you end up removing them or or categorizing people into What what I called the other category before Um, and the main way that we address this is through the the access side. So we might Uh, you know, depending on what the data owner wanted publish a general release That does have the data recategorized like that But also publish a a more restricted version For researchers that might want access to that more detailed information Thank you very much. Um in the interest of Well in the interest of saving time we've got to shut up, but um gen will proceed through we'll go to steve shortly And then I'm going to make an executive decision and say we're going to only have 15 minutes for little lunch And then we got to reconvene because then Levi starts after Little lunch. So steve mckechran Well, I mean, you know studying up steve Is the director of Australian data archive national Australian national university where is responsible for strategy and operations of the archive Has been actively involved in development of application of data archiving and survey research methodology and technologies For over 20 years in Australian university sexter Steve has also had leadership roles professional associations data archiving and Infrastructure including chair of the executive committee of the debacle data documentation initiative pressureer code data data Teach teaching Australian consortium of for social political research Um an international federation of data organizations now when you finish Steve we're figuring out some Structure on the fly When always my approach and Jenny Look at me. Look at me. Um I'll get you to come up straight out to steam or facilitate a conversation in the open discussion before little lunch and we'll Yeah, but we might make up time if we need Uh, yeah, so so we're figuring out a little bit on the fly I have a real strong incentive for this. I'm going to be on a plane at 12 15 So, um, I'm very good at beginning to be able to keep things short An incentive to at least okay Um, we've we moved the order around a little bit because I thought actually Ryan stuff would flow nicely from what Raj was talking about that I mean The iron iris the integrated part of iris and I'm going to say I'm going to talk a bit You know quite a lot about that is we actually have some quite good Separated services in the social sciences That that exists the problem is the integration part. So How do we connect these things up or how do we make choices between systems? How do we Liase between systems so to a couple of the answers to the questions that were raised in the past two presentations was This is not an either or it's an and in almost every case I could do this and I can do that And pushing people through different systems and connecting between systems is really the big challenge We have here and that's sort of the the lead into my presentation as well fundamentally is How do we enable those connections effectively? progressively from manual to some sort of mixed automation and manual semi-automatic processes So, you know, um, Ryan was talking about de-identification Well, if I'm downloading the data, that's the thing I do or all my alternative is to push it into rajah's system if it's got Yeah, different risk We do this in other places as well. That's the sort of story we're going to talk about in this last presentation as well So we're talking. I'm going to talk about metadata a lot. Let's say In this process. I'm also the reason for the all the acronyms to be fair. Okay. So that's my fault That's because saying integrated research infrastructure for social science and the coordination and research data services And social sciences just gets really long very quickly. I just need shortcuts um, okay So I apologize it for but don't apologize for it in a sense. Just say you went up with too many words too quickly Um, so we've talked a bit, you know talked about what is iris But I want to talk just touch on a couple of key points And this comes from from what we included in our original ideas Quite good resources, but they're disconnected and they're fragmented and that's you know Really the story we're looking at here and all the stories we've seen yesterday and in the south in this morning are really about Either, you know raising up the you know the capacity to connect to things or progressively, you know joining joining up that infrastructure fundamentally um, what we've been figuring out along the way though is actually that we need quite a lot more of Things that allow machines to talk to each other fundamentally is the story that you know that we're increasingly finding um, so You know movement of information between systems. We actually have quite a lot of data, but getting to it the access side We do have this history of restricted access And that comes from the sorts of data that we tend to use um, but we also have this um story of Thinking about how to get to things and then how do we describe things we describe things fairly well But we we don't reuse information which allows us to connect systems. Okay So let's say I mean that so we had a couple of you know diagrams that we've sort of flashed up throughout One we've we've tried to think about this in terms of research process So I'm just reiterating things here as we go through The different work packages that we have within iris are really designed to you know connect up different parts of the research process That was that's one view of how we're thinking about things is what do you need to do in different parts of the You know as you work through research activities So that's one framing we've got here is how do we make research easier fundamentally? And this this was in rajah's you know document this morning Michael nicely then talked about actually it's a combination of systems as well So we need systems to talk to each other. So we have curation the cards a vocabulary service That's all just social as the integration services. And we're also interested in doing survey data Managing of survey which is a sort of a workhorse in you know, a lot of the quantitative social sciences collecting and managing survey data It has some parallels as well with a lot of administrative data collection You might be using Government administrative sources as well We can use some of the the same approach as well and then the demonstrate approach that projects are the means for connecting those Okay, so what's I say we knew this from very early on but I say we would need quite a lot of you know Mechanisms for connecting between systems of really the first stages of what we've been doing is is surfacing what those requirements fundamentally are So we need to create Metadata the data is there, but we have to be able to give it meaning and be able to get let the machines to talk to each other Humans can read process and understand imprecise information machines can't That's fundamentally the story that you know that that we have I mean, you know, that's that's an obvious thing to say But what it does actually mean is creation and curation Or in order to effectively operate and interoperate these services is really a lot of the driver for Sort of feeding through what what we're doing So we require systems that can create metadata at the point of creation And that's a little bit of what I'll talk about now A lot of the systems that we have now producing information that allows us to do this if we can pass it to the next system And if we can reuse it And or we might have to curate metadata and data as it passes through systems And that was the sort of infinite thing that ryan was talking about and essentially that michael depends upon to run the geosocial system So i'm revisiting from from michael's, you know, michael's diagrams yesterday We we could see if we can find effective means for describing the data And you know that is essentially what metadata is Um, and we can provide Common vocabularies for the the way we do it. We have things like research vocabularies astral and a couple of other services We've been introducing as part of a part of iris as well Then we have the basis for interoperating But often these things are not there already So either we want to say we want to look at how we can make sure that they get carried through Or we have to think about doing it after the fact So we it's either at the time of creation or the As the process develops and different parts of our systems will need to do that And we've just certainly been discovering and the last couple weeks have been doing some work amongst the teams to say What's kind of missing here? uh And so nicely sort of say that the iran team been working through a metadata model there And you know it touched yesterday upon well, what are those common sets? You know the common standards nick was talking about common standards Really standardized sort of content, you know, is the easiest thing to do or at least comparable content. So So for michael to operate the geosocial service, he needs things that are essentially the same Certainly to do the sort of the the computer, you know mediated reasoning I said a human can do it But they have to do a lot of operations in the meantime The computer cannot do it until you give them sufficient information to do so No, I know we have ml and you know ai site type models that are there But there's only going to get so far. We start to feed them the that you know We have to get away from the garbage inside of it in order to be able to do that so What are the metadata requirements we're looking for and this is you know, really the story of where where we've come to So the sort of things that we're trying to build through in our in the the work package five, which is the survey Data management and work package six, which is the curation services is a combination of you know, some basic requirements for metadata We need some technical and semantic standards. We have to be able to say what a thing means. What is unemployment? What is education? You know these sorts of things to be able to say If I've got a tabular data set, what is in this column? You know, I can say I can give it an aim. I can give it a postcode Um, but if I don't have the the semantic standard to say it's actually a postcode I can probably figure it out if it's got four digits What happens if excel decides to round it to three digits because it doesn't like leading zeros as an example That's the sort of thing you kind of need to be able to do to that machines talk to each other So I need this the technical and I need the semantic. What is you know, um, you know, 2601 actually represent Now that represents the the postcode for anu in this case Or 3437 which is you know gizman in victoria where I spent, you know, I live for someone You know, I know this and I can remember this and you know, you need to present that information effectively You don't so what else do they need to be it needs to be persistent. Can I reuse that from previous work? Um I'm going to ask you to watch closely as they make my way through because it's the same I'm going to tell a story of what looks like persistence that actually isn't You know, can I reuse something and you know and make use of it over and over? If I can build that chain of evidence, I can then say these things are linked to each other It needs to be fine grained Um, so As said, what's the name of this variable? What are the semantics of this variable become important there? um But we also need the course grade and ryan and and and boy tech in the in the last few days have really touched upon this I need the context to understand under which circumstances can I do something or not do something? Um, you know ryan was touching up on sort of populations and sampling uh as as was one tech Do we need to do identify something? You're more like you need to de-identify something that's taken from the census than from a a survey and that's because A survey samples people. You don't know that someone's going to be you know in a list a population At least a census or the case that you were talking about the indigenous professors You know who that list is so you have to do different things. So you need the course and the fine Okay, so we have nicely we have a quite a lot of this Um in in a lot of the social science in you know infrastructure that we have It's just disconnected and often using different sorts of standards. We do try and standardize some of this This is an example the international social survey program, which I I'm involved in the Australian arm of this Um, we have the basic things that we can use to build this fine grained infrastructure So we have a variable a variable is a column in a in a spreadsheet Fundamentally, we can give it a name. We can persist that information and we can say this is the variable that I used in my my asker So that's nice But a variable in a survey and this is good. You know, how do I reuse of information comes from a question? I ask a question of an individual. I might have done it. How many of you have used qualtrics? Um in any that you've done Okay, I can get that information out of qualtrics and I can probably pass it the next system But I don't usually But what if we could why don't we do that? That gives us this word starting to build that sort of provenance change that question is there I also have a response domain. Here's the answer. So I have a question and an answer And the difference between the question and the answer might actually inhibit or support interoperability if I you reuse the same answers Essentially, that's a vocabulary. If I reuse that vocabulary, I've got the same answer to the question Um, so my you know, Ryan was talking about other categories. Who's in the other category or is that list of We won't have a point to hear not going to see it online. Anyway The response domains can be reused nicely. This actually aligns pretty well with Lists that are in up in other sources as well What is your current employment situation? You know in paid work unemployed in education That's something social sciences. We want to compare things pretty regularly We reuse essentially the same the same response categories over and over and we even standardize them into national and international classifications So our research processes support this we generally reuse the things but we don't persist them. We don't always use exactly the same thing so Okay, so we've got the pieces there. We even have a concept. This is essentially employment status We give it a name. We give it a definition But we might we say we've got to be consistent in our names and our definitions and this is some of the hard part So, you know, how I define employment status the abs There isn't this national classification the unemployment statistics that appear in Australia every month are determined by an agreement in 1984 that flows through to the abs And you say and there is a very specific things that are there But I might make it slightly different um So we do have the pieces that we have there and actually what we've been demonstrating in the work We've done so far is we can do a number of these things We actually have been able to in the the process of being working connecting a questionnaire in an archive so getting stuff out of survey services and into registry services um as I is We say so far we've been able to get you know from the using standards This is the means through which we're able to connect content up So we've established a registry based on a tool called collector and using two standards the quick some l standard which is used inside for describing surveys and the dvi life cycle standard which is used for describing Data content in social surveys and and related systems We're able to get content out of archives again using pretty standard systems using spss We could use data sass things that are essentially on every laboratory desktop in every universe in Australia to standardize things they generally describe things in the same way we can move them around Less success with having classifications to vocab. So I talked about classifications So the abs has a standard classification occupations. It's got 6 000 categories in it. There's classification of sex and gender and gender roles uh, not gender roles Sex and gender classification and a sex and gender standard Those are less well presented. So the way in which they're presented matters We can move things in and out, but they're not always available in effective ways And they're not always exposed in effective ways for those who use apis Turns out the api we've been waiting for an api from abs for 10 years We hadn't got it. It turns out it was sitting there hiding behind the scenes But you've still got to find ways of exposing it But let's say a lot of these depend upon and this is the parts that we're increasingly having to do export formats use of standards Moot help, you know How are these things exposed by one system to pass it to the next system and the access points that are there? Preferably apis but there are other models that are there. So we we'd be able to progressively build up So we're our pilots we're getting we've managed to get things in and out of systems into an into a collector service here So we've got this stood up as a bit of a proof of concept We've got three instruments an instrument as a survey a data file is an sbss data file We have over a thousand variables. We can describe it, you know, 1700 columns of data We've got 180 questions in our system We've got pretty good metadata on the system This was all generated already by the systems that we used to collect the data We didn't do a whole lot of curation for these particular questions, but we would have to do this for the census questions Um, we're able to say here's the question. Sorry. That was the question in an instrument Here's the questions that came before and after here's the full description of the question with some standard categories The same questions I used before in my example And then here is the variable that appears in the dataset, which is the column of information the categories it uses So say we've got the pieces that are there. We can describe the content We can connect it up and talk about the pieces that we want So a third view of virus is say, what are those, you know, arrows from left to right? So we're progressively able to build up This is doesn't include the geosocial Which would be, you know, connect into some of these services as well But we have been able to demonstrate through different means some of those pipelines and how they exist And we're looking to progressively automate those But we still need reuse in particular. Now, I talked about, you know, persistence and reuse and sort of consistent use of Information and I kind of flicked over something quickly Here's a question It looks a lot like the question that I had there had there before So which of the following best describes your current situation? That was the same question sample I used before Here's another question. Which of the following best describes your current employment situation? Are these the same thing? The answer is nearly And one one talks about which the following best describes your current situation Which the following best describes your current employment situation We're essentially asking the same thing and using very similar categories, but we don't use exactly the same thing We do the human thing which was we copy and then we modify But the thing for us that we need to do next is connect those up We want to reuse the same thing or where we change things We want to be able to describe that we've changed and that's the bit that we're going on to next Next parts we're going into A registry A collect card and the research vocab is australia to describe to allow you to connect them back to a questionnaire to reuse those questions Those wordings were different and the response categories are different, which means they're not interoperable Can we build that interoperability in as the next part of the work things like question libraries? We're looking at content reuse Exactly, you know that that idea of reusing those questions, you know in effective ways, but also variables and classifications And the likely solution. This is probably some sort of graph You know a link data service plus a provenance model that goes with it and that's really where we're moving into the next stages of work So we have the pieces that are there We're saying we're telling a story of interconnection But progressive as we go through we're starting to think about all How do you actually make effective use of those? You know, we can move things from one system to the next But how do I reuse the things once they're there back into the systems that I have and that's kind of where we're going next And that's our story Hi Steve, um, can you You you've you've mentioned or name drop the vassal thing there And you made a couple of points about you need better Interaction between tools now on the one end of the vessel you've got the rba system You've done stuff to it to make it To make some content there more available for your tool. So what yet? So this is a bit we're actually figuring out now is Is actually what so what does rba is research vocabularies astray at an ardc service? What it does is make particularly good, you know, vocabularies are what it does best So in the in for the what we need to do things like a standard classification Or a list of response categories of what you do west, but it doesn't do the rest of those other pieces The question the variable and so forth It also is the case that an idea that that last point of demonstration The the answers that they're using a specific question Might be slightly different And it you know, so we what we're looking at now is the interaction between the standard Is this you know the list the pick list the things you can pick from and the specific ones that were actually used And what's the best place to locate those and we found that for we're doing some work with the census dictionary For example that it works well for the classifications and for the the standardized things But where we want to make modifications We've got to kind of look at where we host things locally on how do we make those Potentially reusable and probably put them back into rba is not the ideal case there So that's kind of where we again. This is this question of reusability How do we present that back out and describe the line of the other sort of the provenance chain? Thanks, dude. It's really interesting At this stage, do you have any idea of How researchers are going to interact with this? I'm thinking in my discipline There's a great reluctance for one person writing a grammar to say my past tense is exactly the same as your past tense We can use the same category. So if you got any kind of feedback So not yet the the question really is is can we Say because we know Actually in certainly in this area. This is is kind of how researchers work. It is an expectation that you reuse With a bit of a little bit of you know, it's the exactly as you know, so we got a slide similar to as opposed to same as The you know, we do reuse a lot of content though But we do it inherently we say this is the measure that I used Here's the questionnaire that I used but we forget to act. We don't have the systems actually to say This is the relationship between those two things Particularly classifications. We do want to use those because those are the things as that The you know, the geosocial work was talking to you say those are the things that allow us to connect them up and make those comparisons but again, we might We make you know slight changes around the edges So let's say there is a tradition of reuse Um I say with modification So it's really that you know, where do those adaptations occur that can we kind of describe that inheritance? Which is what we expect So the specifics of how that plays out for researcher is probably in those question libraries that I mentioned but You know, I say How do you describe what changes were made? It's going to be interesting question Yeah, I've got a request. Yeah Come in here. Yep. I want you to say this very clear voice. Yes This is Steve mckechran coming to you live from melbourne 3a w Yes, this is Steve mckechran coming to you live from 3a w Oddly that was my mother's favorite station growing it But you got a deadly radio voice And looks forward to yeah, I know that story Um Thank you very much, Steve and sorry I had to we had to rush through that but then You've allowed us 15 minutes to have a conversation now Jenny So Yeah, I mean I say you've kind of got a flavor of Of where we've come to so far and the sorts of pieces that we're looking into there Um, and we you've talked a bit about Let's say this is kind of end of part one Sort of a discussion. We're looking to the future and saying well, what else would you need to bring into the mix there? Um, uh, and what others, you know, what sorts of services what sorts of infrastructure would be You know sort of complimentary to this that doesn't start to appear in this picture now. We we've thought of some Uh things like you know how we support qualitative content So qualitative social research Um, and we have colleagues here at here at melbourne uni who are working actively on some of that work as well Similarly and we're going to say some of those in the room are involved the conversation things like new forms of of data um social media Instead of things type data, uh, so we have colleagues, uh, amanda michael Working on different parts of that um that thinking Um, the third part that we've certainly thought about is And this is a big driver for a lot of the the sort of research that we we tend to see is Is access to data that's not in the systems that we've just described But particularly and this this was touched upon in some of the questions um the uh those linked datasets Um or other government collections that that might be there. How do we enable access to those in an effective ways? So those are sort of initial things that we're thinking of that certainly you know extend the work that's there But you know, we're sort of wondering a little bit about what parts of the story. Haven't we got covered there? Or does that start to tell part of the story for us? So I want to open up the floor to say you know sort of first reflections on there What I mean, what are the other pieces there that you might you know be thinking of and that includes the same We've got a number of indigenous colleagues here Um, certainly looking at indigenous policy. We do see strong overlap certainly, you know occurring there um But you know other spaces as well that might be of interest for us to sort of explore as we go through Nick we've got Nick and Sandra yep A question about um maybe stepping back on and thinking about what kind of government structure long-term government structures are are around Defining the kinds of problems that these data sets and this activity Are about solving and who is at the table in terms of Those decisions and my question because you you know, you're engaging a lot with ATO and How for well funded organizations? um My question would be in regards to closing the gap priority area for Which is all about access to data And I'm wondering what is the thinking around Indigenous led co-design around the problem definition about And where is the space for um The appropriate indigenous bodies to be Defining the data they want to see and that reflects their needs and their requirements and You know, where are we at in these conversations with the government? I won't try and talk to where are we at with government Um, I say what I would say is this is what I think is what so to touch upon the Has an indigenous comments more generally this this is one of the joint problems I think actually is one of those shared you know shared parts of thinking that needs to occur for the has commons It's not specifically an iris problem. You know, we've got part of the story to talk to Um for particular types of processes and particular type of researchers But governance more generally and and I would say reflecting up, you know, so I'm involved a bit in Yeah, obviously an idea in as well Um But reflecting on some of the governance sort of questions that have been coming up These are There are some indigenous specific questions that are relevant there, but actually the governance process we're talking about here in Related communities as well. So I've been you know, I've done some work with the national disability data asset We're similar sorts of stories actually started to go started to come through in terms of thinking about Ownership governance, you know rights to access Certainly have come through in that sort of engagement as well. Um, so I would say I think that is a collective Um requirement that we actually have across the house commons and I would extend that into a little bit of other places as well Things like the health commons, you know The people commons are likely to have those sorts of questions too the other Heart knife and the the questions we got from raj You know for raj's presentation kind of allude to this is there is sort of a governance and and a technology You know discussion that needs to be thought about for I say I really like that the term you know respectful sharing, you know turning around that language Towards how you manage some of what is essentially Government driven processes for access to their stuff for turning things around to actually I actually think though some of those models work for the other way around imagine if government actually had to ask the same Sort of questions for access to your content as opposed to the other way around They're a consumer rather than a you know, they're rather than a producer of data So those sorts of models I think probably have some you know capacity use if we think about the Some of the the government structures we can put in place That hasn't been that work hasn't been done yet And I say I think extending into You know other communities as well I've done work with public housing communities with a disability data asset and the like and I say the the sorts of Questions that are asked and the sorts of considerations people are interested in Come through the interesting part. I think as we go through this is And the challenge that that has to be reconciled is It's a story of I want access to information about me And that's good, you know ownership and control over the information about me is good The problem is I also want access to information about other people And it's those other people whoever they might be Who say have the same thing they want information to access their information But it's that reconciliation the collective and the individual which is you know, there's no answer to that But how do you know? How do you put processes in place that actually laid to reconcile those two things that you know That the parts we find out to work our way through Hi again, Steve just before Sandra's question you you said how do we sort of integrate with other stuff? And I note that in the discussion thing here, it says future requirements for you know, social science infrastructure I mentioned it to do with modeling and so on yesterday, but There are lots of initiatives out there to do more with data That's got nothing to do with social science from their point of view But from our point of view it does and one example I'll use is we saw the The tooling about place names yesterday So when Queensland and hopefully the national governments improve their place names models That could have a positive impact on us But in terms of what we should do to be on board with that is we need to actually come up with Even in concept what we would like to see and then actually present that so when Queensland government says look We're actually rebuilding the place name system. They're doing it right now We should go to them and say Indigenous data network that the has community or whatever This is what we'd love to see now. They're sometimes they're going to say look, we can't do that But for instance if we just said, you know that the government there would understand the dual naming Or multiple name with indigenous names is a big deal But specifically how that's done this community has got the contacts to Indigenous groups and so on that would actually be able to express their wishes And as the mediators people like me in the sort of the technical space between the communities and maybe in this case The government would want to interpret those wishes into an actual proposal because one of the things the government's not going to do is try and Implement your intentions in their model because that's risk for them They're doing stuff for their purposes, but we can probably often add on to that as long as we kind of guarantee We say if you represent indigenous naming in this way in your system, we'd be happy with that They'll take that as that's great. There's no risk for them. They can go and do it. It keeps us happy. So my my I guess proposal here is to look at those other initiatives that are around And actually make concrete proposals to them Even if they really basic starting ones that we think represent our community's interests and and this is an ongoing I guess advocacy task but just Looking at the initiatives that are actually happening and then trying to jump in on them. That's what we need to do Yeah, I flicked It's disappeared off the screen so on my screen it could be pop. Yeah, excellent. Thank you I flicked back to this slide but for the two in response to it with and I'll say now our thinking is exactly that You know, as we know there are some things that are happening in other, you know in You know, you're trying to pull together things from lots of different sources generally speaking and knowing about those I sort of mentioned, you know Use of my standard That's never, you know, that's you know, ideally that's what I want, but that's actually not what happens It's what I want is concorded or crosswalk, you know to be able to connect so Looking at you know situations where I can reuse You know, and I know that I can you know use that information in my system become is equal important So, you know, sort of Michael's got this to interoperate these things. You need to follow all of the chain But at this point here things like Providence Who hosts it? What else we got in their collection methods and so forth I say it's probably the case that I can reuse those in pretty effective ways So spatial temporal the other the other is the kind of the social science the terminology the conceptual side The spatial and the temporal should be able to reuse and I think that holds in a lot of places as well so as I said, you know, it's Looking at common, you know common practices and as you say and feeding in this is our requirements there and be able to say No, how do I how do I align those things effectively? I think is the kind of the model for doing that and then be able to again follow that chain down You know, can I say what the relationship or what the similarity or difference between these two things is? But not trying to do it all ourselves at the same time We know it's like we know there's a lot out there We know that probably it's not going to be our us that does that you know that work But knowing that we can talk to it becomes a key part of that Very good over there. Yeah Thinking about the kind of data that we're producing from a range of projects at at six in brisbane We frequently don't use surveys Because people want to yarn And so, you know, we have these what we we call narrative assessment tools So the the drawings that people collaboratively do as part of an interview process But you know in the course of asking a range of people about the same topic we might get An interview transcript will get a drawing will get notes from the person who was doing the interview So all of this and a recording of the interview frequently as well. So all all of those things Other data I'm wondering if there's there's a scope for a project that is looking at okay Indigenous people were designing a metadata system for this kind of data, which is frequent, you know Ask Blackfella About something he'll tell you a story So, you know, the the idea of of narrative being the the basis of the answer rather than a variable Yep, maybe there's Maybe there's scope for a I look, I I think data project. Yeah. No, I think absolutely again, there are there are It's why I get a little bit concerned about talking about my project as opposed to the has commons as well Which is as I say the I think absolutely Um There's sort of two sides that one is They're all just data Okay, so we've got to be able to store stuff. We've got to be able to talk to stuff We've got to you know figure out what's what's there and you know, so there actually aren't that many types I think, you know, it's image. Yeah, as you say, it's images text audio video Spreadsheets and then there's probably a little bit more Jason, you know, sort of the web, you know the web data standards that are there and you've kind of got to cover So that, you know, be able to access those things is part of that Then there's the semantics of what is there actually in that, you know, in the content that's there as well And how do I how does a particular domain or a particular group make use of that in an effective ways? How do I make that information? You know accessible And and surface that for what are going to be quite diverse You know, um research needs But might not be as diverse as actually, you know, we think that they might be so using natural language Processing to go through and identifying, you know places in a in an interview transcript You know, it is one example that that might be there Now I can see two immediate cases for that which is one is I can identify the place names Which is great for being able to say something about place place, you know Place-based understandings the other applications I can De-identify the places which from the point of view of thinking about data security And and re-identification Is is an immediate, you know application of it as well So I say I think some of the the services that we actually need are probably going to have multiple cases It's how do we take that and be able to think about? What are fairly common data types and data services and tools that we need? And then apply them to particular methodologies of particular domain traditions And requirements that are specific to a particular, you know research community Um, and that's you know, so the the things that we actually do in common. That's the that's the commons as it were And I say and I think absolutely there's a capacity to be thinking about, you know, where we might use those same thing with images We were working with images 10 years ago, we couldn't do this 10 years ago, which was where there was a colleague working on Indigenous football competition, um out in, you know, uh, western New South Wales You could tell you where it was taken could tell you who you know What time was taken what size the file was what the geographic location is the thing I can tell you was what's in the photo You know You can do that now, of course And that's great But I can tell you who's in the photo and that might not be great as well Um, so what again, how do we apply the technology particular techniques that that we're interested in there? So I think absolutely that's going to be the case and we can think about across the sorts of content that we have And then think about well for a given application What's the you know, what's the what's the needs we have there? Um, but in that say we so we thought about certainly some of the access side of that in a related project that we're doing Now what are you know, particularly between you know, the qualitative social research traditions You know and and some you know, the oral history traditions is probably not a big leak You know, so there are certainly some requirements there Uh a quest a comment from online from Julia for local problem solving and local solutions for communities that want to understand their local issues to find solutions Only census or specific administrative data that is geocoded is useful Unless those communities have resources for local surveys This is a problem as otherwise all these conversations are top down Yes and no So as I I made my start doing community surveys in particular parts of the local community Yes, you need the resources to do those things But again, it's not an either or sort of question So be able to work downwards from those, you know, large collections is important to provide some part of the equation Being able to provide access to Local like I mean say the work I was doing was training local people to do surveys of their own community with some with some supported resources Um, but those tools are becoming increasingly active. The the survey tools that I was discussing Uh are not free, but they're not expensive. They're getting their desktop tools at the university level They're relatively cheap for a particular community survey monkeys, you know, google docs Uh, sorry google forms can do these sorts of things So the technology is getting less and less expensive than they were the capability is different Um The the yes side of that question is Once you start getting to those low levels That information starts becoming again, we get back into confidential deli and privacy issues as well Um That that become part of the equation too. So to say it's always this trade-off Um, and there are I say some really good work going in and certainly indigenous communities to to think about how you do that Uh Yaru up in uh outside of broom, uh, for example, we've been doing work with Andalaqua And just out off Groot Island Who are working on these in their communities to do addressing because sometimes the high level data just isn't isn't worthwhile Frankly, it's not sufficient And so you do want to do some local stuff Again, I think it's got it getting to an end rather than an or needs to continue to be part of that story Shall we kick back in ladies and gentlemen? We're still in the process of making up time. Um Hello, Anika so Levi and Kristen Levi Murray and Dr. Kristen Smith are going to Do their presentation or improving indigenous research capability stream one social architecture Now Levi are you related to arnold? bumper Yeah Yeah, him and I played football together Yeah, um Levi is an Aboriginal man of walker walker and cubby cubby descent Um, what am I? Lee Sorry, Levi they go and leave on then they've pronounced you Murray Um Well, Levi isn't Murray Um the expert on health strategist was over 16 years of cross-sector experience in health and education Worked as a that dynamic range of pro in a dynamic range of primary healthcare settings throughout Australia Error error medical youth and adolescent forensic health rule and remote health health age care and disability committed to perpetual independence growth evolution of healthcare infrastructure for our first nations people and their respective communities throughout the global COVID pandemic Levi delivered senior clinical leadership and advocacy to victoria's Aboriginal community controlled health sector and nationally as an executive manager of clinical excellence Murray Murray Is it Levi or Murray? What's going on here? Hey Yeah I don't know what you're smoking when you put this together Johnny, but the expertise in both public population health was central to develop in developing COVID positive pathways for Aboriginal and Torres Strait Islander people and the co-development of new clinical models of care for COVID-19 vaccinations pathology and diagnostics April 22 as appointed a strategic manager of indigenous data with the indigenous studies unit at Melbourne school of population and global health overseeing the house the research data commons and indigenous research capability program project Murray will adjunctively can co-continue to the relaunching of my god Nadia Yep, and you know what it is and the development of their five-year strategic plan and dr. Kristen spear Is a research director of indigenous studies unit in the center of health equity at the melbourne school of population and global health the university of melbourne Kristen holds phd a medical anthropology master social science international development From rmit and a bachelor of arts with distinction from deacon university her interdisciplinary work traverses the fields of medical anthropology epi epi deamoled the epi Yeah, epi deamology human geography public health and health promotion Kristen's unique contribution to health research include her innovative work in Australian indigenous medical anthropology policy analysis data analysis complex multi-disciplinary theoretical and methodological methodological development And that'll do I think So narrow one and all my name is Levi. That was a very long winded and very complicated by an intro So apologies for that. Um, I would like to pay my respect by acknowledging The burundry people of whose land and country that We're joining from here today I'd also like to extend that to all Aboriginal and Torres Strait Island people joining both here in the room in those virtually Um and extend that also to the many partners of the many Communities in which we work with throughout the multitude of programs and projects that we function within the indigenous studies Thanks, Levi. So we're also going to give a bit of an overview of the improving indigenous research capabilities project As although a lot of the other projects have had quite good overviews and details We're going to give a very brief update on some of the activities in the first part of the project and then following us Sandra Len and nick will be talking about two of the other streams of work so The project partners and like the other projects you've been hearing about over yesterday and today We have a significant number of research partners organizations. Some of the key ones are named here But there are a number of other organizations who we are working with as well Again, this is just some of the key team project Project team members There are many more people working on different activities in sub activities around the project This is just a representation of some of the key people working within this project So just to give you a little bit of a background on I guess initially the indigenous data network itself so for many years Personally myself. I've been working with Professor Marcia Langton for a bit over a decade So we've been working on a range of different research projects both being anthropologists Mainly across the northern jurisdictions of Australia from Cape York to northern territory And more recently in the last four or five years We've done a lot of work up in the east kimberley and we've started doing some more work In central australia across the mpy region so I guess probably for quite some time during our research. We always work with community controlled organizations And one of the key things that we always found frustrating and that the communities that we were working with found Extraordinarily frustrating was the lack of access that they had to data So when i'm talking about that it's primarily we're thinking about government data Assets that were about them that often they are sending off data So particularly if we're thinking about health services or any other range of services in different communities across australia Yet they have very little access of that data coming back to them And some some organizations are working with this data very well and collecting their own data But other organizations who have far less resources and capacity to work with that found it extraordinarily frustrating So as I mean a lot of the researchers in the room and listening in online would understand it even for Researchers in different fields. It's very difficult to get access to government data And government data that is meaningful for particular community groups for particular issues So on the back of that marcia in a lot of her working Um Decided with a group of us to put together a symposium looking at indigenous data sovereignty. So that was back in 2017 We had about 200 delegates that we invited from all over australia from government from the Different peaks across australia from community controlled organizations But also a series of different experts across the university sector In all majority of universities across australia So that symposium was co-hosted by iaccess the australian national university and the university of melbourne And so one of the key I guess outputs from from that symposium was the call for a develop developing a network of interested parties in promoting indigenous data sovereignty across australia. So really that's where the indigenous data network itself came from So I guess fast forward to today The idn really is now a national network and it really aims to provide this collective base To support and drive better data outcomes for Aboriginal and Torres Strait Islander people across australia It's made up of a series of you know community controlled organizations There's indigenous privately owned businesses There's a lot of people in academics across the sector predominantly Aboriginal Academics but also other non-indigenous academics who are working in different areas that are focusing on indigenous studies issues So one of the primary aims of the indigenous data network is to support community controlled organizations to collect manage and utilize data assets relevant to their needs and aspirations and to really think about how we can collectively promote indigenous data empowerment Yeah So some of the core issues that the indigenous uh, um improving indigenous research capability project Set out to to answer and review is indigenous research data One of the things that we knew is that it exists in a wide variety of formats um And in a wide variety of context, but these assets are ultimately scattered and inadequately cataloged There's many unidentified yet highly significant indigenous data assets located across the nation, which many are orphaned Or held within five broad communities of indigenous data research custodians Thus including original Torres Strait Islander communities and their organizations universities researchers and other organizations The gland sector which is galleries libraries archives and museums Commonwealth state territory local governments and The fifth one the private sector So As many people have already been talking about over the last day or so What what we do know is that you know, many of these cost data custodians are really siloed So there's not a great deal of integration between them And they also have really widely varying data science practices and capabilities And so we're not just talking about the community controlled sector We're also talking about you know different data holdings within Commonwealth state and territory and local governments There will be different ways that data is managed multiple systems And it's really very complex for you know, just average researcher, let alone Say a data officer in a community controlled organization to access Also, there's there's Really been an increasing need for a variety of solutions for data storage and management across all sectors and all of these different data custodians So for example, you know Some of the communities that we work with have really large holdings of data whether that be video data Increasingly so which is as we all know really large files that they're holding on hard drives some of them in remote communities and often in areas where they are really at risk of Of damage from whether it be during wet season the cyclones coming through or for fire and For one one community that we're working with They really have little ability to upload it to the cloud because the lack of internet service or the limited internet service that they have in that location So thinking through solutions for really varied problems such as this is okay So how do we get you know, however many well a quarter of of a Pedabyte of data onto the cloud from a remote community So I mean there are a number of different ways that you can go about that But what actually is going to be the best solution for that community who is really concerned about ensuring the safety of that data The the longevity of that data and ensuring that they safeguard particular elements of that data in the way that they culturally wish to either share or keep themselves So that's just one of multiple Issues that are arising in different ways for communities across Australia Um, there's also a number of different indigenous data custodians who are calling for advice and practical solutions Of how to actually embed indigenous data governance and indigenous data sovereignty principles So at the indigenous data network, and I know probably a lot of other organizations out there We it's a common request that we get from particularly government or government agencies of saying, okay, we're custodians of all of this data We want to implement indigenous data governance principles. How do we do that? And so in this project We're really wanting to think through some practical ways of different agencies and organizations And how they can start thinking through embedding these principles within their data holdings And in the way they collect the way they analyze it the way they share it so across the entire data life cycle Um, there is also, you know a variety of other complex issues emerging Regarding data access for indigenous people and that's already come up a couple of times across the last few days people talking about I guess we'll go back again to government data custodians holding access to data and often there's gatekeepers who Do not provide access where access should be granted. So there's a a seriously Large problem looming there and how we deal with that is something that the that this project is trying to start to build solutions for So again, sorry So again, some of the key questions addressed by the project is how do we embed and support sustained indigenous data governance across Australian data infrastructure? How might individuals organizations and government use best practice models to ensure Good governance and what incentives might assist the prevention and use of data into the future? What types of authentication practices processes embedded in IT infrastructures could appropriately protect data on behalf of individuals and their communities? What are practical ways to apply fair and care principles? Are there missing pieces to the puzzle that need further examination? How might we legislate regulate data governance to benefit Aboriginal and Torres Strait Islander data within the jurisdictions across Australia? And how can indigenous data data assets be improved by embedding indigenous knowledges in new and innovative ways? So really it's it we're not being ambitious at all in this project thinking we're going to solve all these things Obviously, that's not what we're trying to do. So this project is about starting to build those foundations That nobody actually has the answers to any of these particularly in an Australian context Where there is such a wide variety of different needs and understandings and uses So the project itself really is split into three streams of work Of which each of these streams have a series of I guess what you would like to call mini projects embedded within them So the first stream as you can see here is the social architecture work The second stream being the technical architecture work and the third being the building of national data assets So of these three streams, um, they're really we've designed them to collectively to lead to improved indigenous research data practices across Australia So although this project is relative, you know, when we were planning it is relatively short So 18 months from start to finish by you know, I again quite an ambitious task that we're embarking upon We really see this as more of you know, maybe a 10 to 20 year prospect of actually doing this properly So what we're doing in this phase of the project is really trying to get some solid foundations to work forward in the future So stream one which we'll go into a little bit more detail as we go along We're just we're just going to give a brief over really brief overview of all of the three streams And the the next presentation with our technical team are going to give more details about about stream two and three But stream one really is made up of three sort of minor projects or activities the first being national scoping and engagement of indigenous research communities and place based application of indigenous data governance principles So that this activity of work is well underway It's consisting of a range of things including we're doing a major literature review scoping review of both international and national literature Relevant to indigenous data governance. So also issues around indigenous data sovereignty That's being drafted at the end of This phase of the project. So end of june We'll be producing a working paper which outlines the findings of each of that these activities and gives a bit of an overview of the work that we've done So activity 1.2, which we'll also talk to it in a bit more detail as we go along We had a national leaders of indigenous data governance and sovereignty roundtable that we held Precisely here in this room over two days in june the 9th and 10th So we'll give a little bit more information about some of the really fascinating sessions that were held there and discussions With a variety of different groups from across australia There's also a link in in one of our slides later If you weren't there if you're interested in catching up on that the majority of the sessions we recorded on video and available to watch in the future So the third activity is actually one that's going to take us quite a lot longer than the 18 months to do Is looking at place-based case studies of aboriginal and Torres Strait Islander data governance principles So in this phase the last 12 months of the the project we've been Doing consultations with a range of different communities across australia Trying to get a grasp of some of the key issues where different communities are doing very well at embedding indigenous data Governance into their practices in their work And looking also at some of the barriers and needs going forward So we're we're working at the moment with Three communities, but it's going to as as we've all been talking about again over the last few days. This isn't a short-term process There is a high level of sensitivity around the data that's being collected by a lot of communities and we've got to be respectful It's not a quick process So we're we're making steps through that and we we do have a good deal of work yet to be done But the start has been made on that as well Talk briefly just The overview of string to yeah um, so string to the technical architecture like christin said Nick and some of our partners will do a demonstration talk through that in a little bit more depth but Brief overview is working to establish the foundations of an aboriginal Torres Strait Islander research data commons Exploring repository services to ensure that the preservation of data assets Catalog services to make them discoverable in the metadata generation to improve asset description and facilitate data linkage and aggregation So key activities again breaking down in there. There's three of them 2.1 is the aboriginal Torres Strait Islander Research data engagement scoping activity. So that works been currently underway and what you'll notice about A lot of the activities and a lot of the projects across The three streams is that they are all interconnected. So One might proceed the other but it doesn't necessarily mean that that work stops at the at that point of the project activity 2.2 is national aboriginal Torres Strait Islander research data catalog An activity 2.3 is the research data capability building for aboriginal Torres Strait Islander organizations and researchers So really going to understand him again one of our primary focuses is about improving that capacity Not just from a research perspective, but obviously from an operational and a workforce perspective as well So then just briefly the third stream so that as you can note there's three activities in each stream It really aims to build the foundations for a set of core national indigenous data assets focusing on the development of an indigenous locife framework, which parallels the ABS hierarchy and can be applied to any data, which includes location information So as we've noted more detail will be spoken about these two streams of work as we go along Um, and just think I want one thing I wanted to note So the activity 3.1 looking at vocab development and metadata labelling for traditional knowledge and biocultural labels A little bit of well that has been worked on for the last 12 months But some of the development work from the technical perspective We have realized that that actually needs to go back up into the social social architecture because Although there's been a lot of international work and there's developing work in Australia There's still a lot of governance issues that need to be worked through with that In the second quarter of this year We will be convening an in-house workshop of a number of different Stakeholders across Australia specifically to look at TK and BC labels and notices and looking how they can practically be applied, but also thinking more broadly about How they can be usefully Used in Australia and really looking at specific communities again There's such a very context across Australia. We can't assume that one size fits all So that will be happening in the second quarter of this year as well Yeah As we mentioned last year in June we convened a round table and we had over 180 participants come from around the country Again across many different sectors And including the Aboriginal community control sector Three of our three of our research partners Did present some of their key concerns and some of their key interests In the realms of data and research and also of particular significance data governance and data sovereignty within the context of First Peoples So the internet's data around table are provided and for all of those attendees to Collaboratively explore and discuss those practical solutions key barriers and models of governance for a shared benefit The presenters and attendees discussed and shared international best practice data science data governance and Aboriginal and Torres Strait Islander knowledges and world views So just to give you a better understanding of the key sessions and some of the issues raised As I said, here's the link to the Vimeo if you want to go and actually look at those sessions yourselves We also published in the pursuit, which is the university's Monthly magazine online We've got an article in there that we wrote after this symposium sort of Highlighting some of the key issues. I should have given you the link, but we can send it through if you want to send us an email So for that the first session was looking at Indigenous data governance and sovereignty in action in a community controlled organization. So that was a fabulous presentation by the Aboriginal and Torres Strait Islander community health service So three of the people at that health service. So there was Raymond Broncker Jared Alexson and Jonathan Leach gave a really good presentation of the journey that they've been on Working substantively with QUT Becky Cook over here and a few other people have been really involved in that work And I'm just trying to see where John Willis is around here somewhere John could talk to us Extensively about that that as well But if you want to have a look at that session, it's really interesting the journey that they have gone on To try and start working with their data utilizing it in particular ways And the way that they are managing it and look it's a really interesting and I guess I would say Australian best practice example of what to look for in Indigenous data governance and sovereignty practice across Australia And then next up we had the Indigenous data sovereignty into Indigenous data Indigenous data governance into Indigenous data sovereignty presentation from the Mayam Nairi Wengara Collective Which is a series of three really interesting presentations from members of that group It was then followed on by The National Indigenous Data Catalog and Practical Applications of Indigenous Data Governance By members of our project team. So Sandra and Len and Nick We gave that presentation We also had a session on the traditional knowledge and biocultural labels from Stephanie Von Gavel at Syro And Professor Rachel Ankeny from the University of Adelaide. It's Rachel here today. No So they gave a really interesting update on the work that they're doing going forward and some of the key issues that they were coming across We also had a Session on co-designing an Indigenous data roadmap and building capability by Professor Peter Anderson Professor Matt Belgard and Professor Professor Kerry Minguson and Becky Cook who have already pointed out over here from QUT Following that we had a session on working with international standards the European Data Protection Act so that was from Dr. Chris Wilson who was an expert in the area Has done his PhD in cyber security over there and gave a really fascinating talk Looking at how I guess what is happening in Europe? We really should be thinking about significantly in Australia to improve the way we deal with some of these issues um following that we had a fascinating Talk from Michael ed at the University of Queensland from the Anthropology Museum Who Jenny already referred to yesterday Starting to think through some of the ways that I guess as I mentioned before some of the gatekeeping occurs around materials data and objects So Michael went through a whole series of ways where I guess gatekeepers were keeping Indigenous people out from accessing In particular photographs of their ancestors. So if you're interested in checking that out that also that session is online I would highly recommend it Following that we had a presentation from Steve McKerrin who has left the building now But he was talking about starting to think through some of the federated Data models that could be really useful when thinking through data holdings for Indigenous Indigenous data per se So as I've already mentioned a lot of the the communities who hold their own data really don't want it open for everyone so thinking at least how can we make data findable so The catalog team the team who've been working on the catalog will talk about that in a little more detail um Then following that it was another great session. I'd highly recommend checking out if you weren't there um that was delivered by empowered communities and a member of the Aboriginal and Torres Strait Islander community health services looking specifically and and talking about the data that the communities need so their communities themselves um Calling out I guess to governments that they've been looking for access to data for a very long time And I mean I guess there's the side issue with that that has been happening with the new legislation that came through last year And the data code for the data transparency and availability act um From the IDN's perspective, we've got significant concerns about The further barriers potentially that they that will put in place for communities trying to access their own data So I believe I do you want to tell talk a little bit about some of the conversations and the groups that you've been involved with in that context yeah, so part of that work of I've been sitting on uh the deputy secretaries uh data champions working group. So the Australian public sector is currently Moving towards developing designing a framework and how it is that they actually manage those holdings particularly those of indigenous peoples With data sovereignty and governance embedded throughout every every part of that the data life cycle from creation from conception to creation and all the way through to management Like rosen was saying with obviously the introduction of the data and that passing through one of our main concerns is the I guess the the potential for that to bottle that Information and put a further strangulation for community to be able to really access critical information and to be able to make determination and Do decision making and all that kind of stuff? um, so yeah, we've been largely advocating around that and so that works progressive and ongoing and um, so we'll continue that work as part of an advisory group Over the coming year But I guess one of the other things that we're doing is we're essentially focused on improving uh research capability And there was a lot of conversation yesterday um Around like research and I guess that's what everyone's essentially here for today I will kind of Go off on a tangent here, but I did want to make a few comments about yesterday. We were talking about research. Um language matters and I think All the way through. Yes, I found it really interesting. We We definitely do identify like researchers as a very particular cohort within community and throughout our society as a whole But you know at the end of the day people like everyone's people And you know as human beings we're born with a very innate sense of um ability to you know, functionally filter through even data Like a whole world is built up on data. Everything is about that um, and so what When we're looking at, you know, indigenous data governance sovereignty when we talk about long-term infrastructure and you know reform and all of these things that um Our high priorities and concerns for us all We can't have that without changing the language at the same time so you know When we talk about research community that in itself is exclusive for Aboriginal and Torres Strait Islander people It ignores the fact that Aboriginal and Torres Strait Islander people their knowledges their practices culture and laws Um far exceed any of the research in more contemporary times because it really is you know a couple hundred years is very contemporary so, um One of the things that we're also doing at the moment is developing Uh further training and trying to roll that out low cost and you know, ultimately free Um, we are participating as well next week in the ARDC summer skill school Um, as Jenny mentioned, there's been a specific part that we've designed for Aboriginal community control and so the idea there is to Basically change the language and we want to be able to create space and opportunity For for them to come into that and ensure them that they have a seat at the table because Even now when we're talking about innovation everything there's not really a clear idea of where people sit particularly Aboriginal and Torres Strait Islander people so Thanks Levi And really that's that's the main part of our talk today. We've tried to keep it short I just I guess in in concluding look as I said initially We really just just are at the start of a really long-term process We're going to be planning out and I guess learning from our findings thinking about okay What's going to happen in the next year? What's going to happen in the next five years and beyond that? And ultimately this is a collective Project, it's can't be something that's just done by one group or one Organization or one discipline if we're thinking about research It really has to be a collaborative effort across Australia So if anyone wants to be is interested in being involved, please do reach out We're based Administratively in the indigenous studies unit, which is actually now not the Center for Health Equity with a Namda Center for Aboriginal and Torres Strait Islander health and well-being That was launched Tuesday Again here in this room, but please do reach out. We're really interested in hearing about Different ideas from different people across Australia is as I said, it really has to be that collective effort But thank you very much Any questions? Any questions? You're accessing information that allows for a recognition of kinship decisions you're making As in not one person holds all the story Yeah, yeah That's a very good question And I think that's very much centered around like that exploratory part of it is understanding the the incredibly The incredibly I guess nuanced elements of not just governance, but also policy and also Respect and rationality as well. So, yeah, any other questions? Yes Thank you My questions about something that Louva said was around the data champions the depth sex data champions. Forgive me because I Know nothing about it, but I'd love you just to say you talked about you sitting on that around that table And what their kind of agenda is and does it include, you know The entire public service or are they sort of select departments? It'd be interesting to know what that forum is doing Yeah, so I'm sitting on that. I'm sitting on the working group of them. So Professor Langton sits in the actual Committee above and the working group is actually sitting there and it's got a quorum of equal indigenous Participants and non from the APS sector Central focus on looking at reforming. So the framework the framework that they want to introduce it. It's theoretical But again, that'll probably they want to introduce that over a two to three year period So that is currently being socialized in March And so if approved and that that's moved in that ultimately would probably start at a very smaller portion of the APS Of particularly interested to probably be the likes of the Australian Institute of Health and Welfare Land water environment some of those The more central pressed ones and the ones that have been identified as priority areas from Aboriginal and Torres Strait Islander people Because I don't see a question enough or yeah Are you guys having any? I won't say blockades, but if you've got are you having trouble with any Communities, I know you guys are trying to do more on ground and in community talks Are you having any communities and or elders that are? Refusing to want to give them information because they want to keep it within community or within just their section of culture And if so have you got strategies around that? Are they working? Well, we're not necessarily going out to seek people's data It's really more about supporting them to use data different groups to use data in whatever way that they want to do I mean if for example, there is need for Repositories, which we're hearing a lot We will be in that role of trying to support and think through best case solutions and scenarios for each individual group But a lot of this I think in this particular initial work is starting to find out Is starting to find out what are some of the variety of issues that different community groups are having So we're not having any specific issues in getting access to whether it be data or communities We work with a broad range Of communities across Australia through a lot of our research works through the IDN and other work that we're doing But it's more about starting to document some of the different challenges different groups are having Yeah, is that is that yeah, and I guess to kind of further add on to that One of the things is is we haven't actually kind of set those priorities So everything that the IDN has done has ultimately come from, you know, very considered Consultation over a very prolific period of time. We're talking over the space of a decade, right? And so I think like, you know, thus far into this project and where we are currently at We've not experienced like, you know, I guess any sort of kickback or that because ultimately we're not determining that, right? So what we're actually doing is asking what is the priority? So it is very explorative and I think in that space, um, it's very I think it's a really kind space. So what we're being met with Is not just really good reciprocity from community particularly key people of interest and elders and Any of those stakeholders Yeah, it's it's actually quite the opposite We're hoping we can't actually Keep up with the demand of any engagement from people, which is good, right? It's a really good. It's a really good problem to have so Hello, thank you very much for your presentation. Um, I had a question you talked about One of your projects wanted to implement indigenous data governance I was wondering how does that look in practice or Is is the work you're doing trying to understand what it means? Oh, yeah, how would you? Yeah, that is a fair question. Um, Yeah, wouldn't we all like to know but I guess that's it. So one of the things is understanding that, right? And I guess it's been mentioned throughout the last two days Just how incredibly nuanced but also very idiosyncratic, right? Would like to reiterate the point that uncle gran made it's just like every single Aboriginal community might have Reoccurring themes reoccurring like priorities and whatnot But ultimately each of those communities what happens and occurs in there those events are distinctly theirs And I guess the solutions and the way that they Have not not not only done governance for millennia, but how they envision doing governance into the future is Yet to be determined and to be understood so super explorative And I guess also the next presentation will be talking a little bit about how they have created some tooling around the care principles that I'm sure everyone will be quite interested in so Coming up next From the google united nation governance principles that would Be something too that you would apply Um, and I think from memory the consensus I already had to Participate or even making sure that minority voices in the context of a group of Aboriginal people so that all the voices are heard And the other thing is that the complex like in a in a western project Management Context they talk about scope time quality cost Risk as an example and then the the concept of Aboriginal time versus western time and all that sort of stuff the same things Have got to apply when you apply western governance into a trauma Traumatized space that sort of stuff, which is what you're across anyway But you're across it, but the others do have to do it as well Thank you very much. That was no other questions Thank you now look at Look at Sandra sitting there all champion at the bit ready to go I haven't had any dirty since your talk spoke with me this morning So you're going to take us through with Sandra silcott lenn smith Harrison Ford And nicolas car So Sandra as you come up. I'll just give you a bit of a rundown Silcott member of the idn technical reference group experienced talented practical ict professional Supported the development of major idn projects indigenous research capability funded by ardc Sandra brings experience identifying users needs in the context of implementing Agreed solutions and empowering colleagues Employee skillset including systems administration database development And data conversation or large research projects including arc founders and survivors project Dr. Len smith Harrison Ford writers of the lost ark What was the other one you were in the The spice one Star Wars Now dr. Len smith lead lead of the idn technical reference group is nationally and internationally renowned demographer With an extensive background of research focusing on Aboriginal torsion on the people dr. Smith was able to reconstitute the Aboriginal population that Was not recorded in the protection board records and demonstrate how the victorian Aboriginal community recovered to reach its current strength Of over 30 000 people who identified the last Commonwealth census Dr. Smith is also a member of the idn steering committee Now dr. Nicola's car, that's you There's a car You can say that that that works in English Yeah, but what is it? Well, like many people here, it's comes from a different corporate background. So yeah Nicholas that's better Couple of days that's some rather Nicholas is internationally recognized as a thought leader and expert in data architecture data systems design link data semantics semantics Technologies design was designed lead nationally significant innovative link data And semantic web projects including location index logitudes Trudinal spine or government functions Also outstanding academic achievements including phd engineering from university of melbourne bachelor of engineering honors for university of sydney Active as standards editor w3c's data but data set exchange working group and open geo geospatial consortiums geospatial working group and i'll leave it at that Okay Thanks Grant. I'll do a quick introduction Well, I'd like to acknowledge the traditional owners of the land on which we meet and The lands on which people are Coming into us from remote relocations Uh, and also acknowledge the traditional owners of all the unseated land around this nation And acknowledge the fact that we see solutions to that lack of seating through the process of treaty voice and truth telling Which we hope is uh, certainly be embarked on like also to uh, convey apologies from our colleague and Leader sam provost who's had to go home for sad family reasons Uh, we'll try to fill in the gap that he left So, uh, nick, I think we'll pass it over to you to start talking about our technical stuff Hi everyone and apparently 20 minutes ago all the state and territory Ministers uh, chief ministers and the prime minister signed on to an in statement intent about the voice. So Things are moving there So this is about the streams two and three as as mentioned in the previous talk Everything we do is suffused with issues to do with governance and engagement But luckily day to day That's not what we're doing here. Uh, we are dependent on Relations to community and so on From the wider idn but day to day We're doing the technical work of implementing systems and models and things like that But I hope throughout the presentation to indicate where the governance stuff comes in and the relations come in And I'll pass to Len and Sandra to get me out of hot water every now and then Okay, so very briefly Um, we're going to overview our technical approach And in fact, it occurs to us looking at our slides that we're going to do a less technical presentation than some of the other ones You've seen in the last couple of days. So that's good um We're going to talk about our data integration framework concepts very briefly Um, and then we're going to go through the actual progress report of where this part of the indigenous data network work is up to So that's the indigenous data catalog project progress We're going to talk very briefly about sam's an uk study and we're going to quickly mention at least some of our plans for the future Now, uh part of the work that we're doing is about Putting more of our information in more accessible and um and readable forms Um, so a lot of the stuff that's presented here today in our work will end up on the indigenous data catalog project website And what we're aiming for there is to present not the stuff that we've done sort of, you know, we did ab and c but more You know, you may have heard of us from and this is what you might be interested kind of think things that we expect and hope people will ask as questions and then The way we answer that will be sort of, you know, indifferent sometimes they'll say how do we how do I make a catalog for my Often indigenous data. We might have a thing about that but we're going to try and work out the big questions that Um that people have of us And then sort of place those in there. It's very hard to read on there But you know finding indigenous data characterizing indigenous data, etc, etc They're not quite ready yet, but should be soon Okay, so the overall technical approach and I think I Now Jenny, um, I think we might have dropped a slide in there somewhere. There's something missing, but anyway, we'll survive Is it is this is not this is an online one. Anyway, anyway, we'll we'll survive Okay So one thing that we're doing a general approach that applies to both the cataloging component that we're working on and the handling orphaned data is to make implicit knowledge explicit So imagine you've got a table of information Um, you know, it could be statistics or who knows what? Sometimes you have implicit information needed to understand that table of data And part of the mechanisms that we use and the style of data that we handle Calls out those relations. So in that first row, there's three cells there and there's relations between them And we're actually describing those in a particular way in a technical way But we're taking implicit researcher and hopefully indigenous knowledge and making it explicit That's part of the preservation process of that information and it makes it On the surface easier to understand We're directly reusing defined concepts and this is where I'm happy that Steve and others have done a technical presentation before us Steve talked about vocabularies and those sorts of things. So where those exist. We're employing those They don't exist for everything. So we have to define some things that are of interest to us that don't have Definitions elsewhere and this comes down to roles and responsibilities Information from the indigenous community that hasn't yet been represented in forms that machines can read and so on. That's part of what we're doing Okay, and the last point here is that yes We are cataloging things and we're calculating things. We are actually giving The the score of certain things ago. We're trying to score things We're trying to implement fair and care scores. Are we going to get it right? Eventually, obviously not straight away. This is difficult stuff We'll give it a go and we'll see how we go. But without metrics government really doesn't respond very nicely. So we do sometimes have to calculate these things Okay, now our general Data integration framework looks a bit like this We may have a block of indigenous census data and we can do the sorts of Things that best practice would have you do regarding that data defining the questions defining the metrics and so on But for wider integration It's very important that we define the reference data around that and again as steve mentioned in the previous presentation Use reference data that other people have access to like vocabaries models, etc So that we can join our data to other people's data. How does this occur? well the places that we use and and Kristen mentioned that one of our charges is to develop Places of importance and utility to indigenous data And if we had our data referenced against such places We might be able to spatially integrate those places with non indigenous data Such as Australian statistical geographies natural resource areas, etc. So this is a well known Type of thing But just going back a slide actually we're not just talking about the places We're talking about the methods that are used the observed properties the agents That's people and organizations a whole range of different parts of the data that we want to Set up as reference data sets to allow us to join our data to other data So here's another one The observed properties the stuff that we're actually all the dimensions as statisticians would have them the dimensions that we're using They might be custom in our domain. They might be you know Something to do with indigenous identity which isn't present in normal census data or isn't present in non indigenous Um statistics, but we want um the the dimensions of our data to be available for reuse And also we want to join where we do have data in common with non indigenous stuff We want we want to join on those points And then finally what does this let us do well it allows us to join our data through reference data sets Which may be vocabaries, but they might be very big complex things like enormous databases of spatial information Um, we can join our data to other data Okay, so that was the kind of conceptual sort of where we're at now specific um progress in the cataloging project that we're involved in i'm going to talk very briefly about the The catalog standard profile that we've been implementing some vocabaries an agents database that we're building reference spatial data sets scoring and then Not number six. I've talked a little bit about number six already Okay, now i'm not expecting you all to be able to read and take in all of that But what that picture is showing is that in our catalog profile and what we mean by that is there are Standard mechanisms to catalog data out there and we have specialized one of those mechanisms to specially cater for Indigenous concerns, which i'll get to in a minute But in doing this we haven't just written down A standard or a variant of a standard We've provided a whole pile of resources to help people use our specialized version of this including things like mappings from our system to Schema.org which can be used with ro crates mappings from what we're doing to the commonwealth's 26 data elements that they want to collect and things like that So there's a whole pile of resources some of those resources are definitional some of them are guides Some of them are software tools, but they all together make a bundle of tooling and bits and pieces which help you with This catalog profile and the profile is used to characterize data out there If you've got a data set and it's indigenous and you want to characterize that indigenousness This is our proposal to do that using mostly a normal cataloging model but with extensions This model has a formal backing. It's not just a loosely defined thing in a word document Take note prime minister in cabinet This is in fact a formal model that we've mostly inherited from other standards and then slightly extended Again, you don't really need to see the details But what you're seeing there is a formal diagram and all of the documents and tooling and everything that we've provided as part of our Profile is derived from this formal model and the formal model has calculable ends So we can we can use that model to and test whether certain Data according to that model will or won't answer certain questions some of the tooling that's part of the profile is a Well, one of the tools is a metadata creation tool So there are lots of tools to create the metadata we've seen describer and others for ro creates information This tool is tuned to create metadata for The representation of indigenousness or indigeneity now We don't expect people to sit around all day and using this tool We're not asking people out there if you've got 10,000 data sets to type them all through this tool But the tool is there to check Uh, if I do have a data set and I put the information here, what does that look like? What scores do I get for that? It's a testing tool it tests the model and it is there for the creation of new metadata If that's what you want to do, but in most cases what we're hoping to do is to find data that's already already described and to Interpret that description in in the in the form of our profile But nevertheless, there is a tool that's online and you can have a go with that tool Now behind that tool and behind our integrated data generally are a series of vocabularies As steve has mentioned you need these vocabularies to be certain about values that you're applying to things Some of the the the the the vocabularies are what you'd expect, you know The Indigeneity of data and by the way, we haven't made most of these up We've mostly taken them from from literature and elsewhere But you know the way in which data is indigenous for instance That is a a vocabulary that exists in literature We've pinched that but there are some surprising ones in there things like The roles of catalogs with respect to data. This is not an indigenous concern. This is a universal concern I've got a thing it's in a catalog. What is the role of that catalog with respect to this? Is it the first paid places was ever published? Is it a harvested version? Is it the primary etc? Those are normal cataloging concerns which This project has brought to the fore and we are Trying to solve those problems those representation issues and hope to push that information back from the data network to non-indigenous cataloging systems So that is a benefit that non-indigenous people will get from this indigenous work And that's because there are certain issues to do with Authority and governance that we can't answer about data unless we know where it was first cataloged and when and For the indigenous purposes that that the systems required, but it does also solve non-indigenous issues Another major part of our Cataloging project that we didn't fully anticipate when we started but we do now is the The need to have a listing of agents. So these are people and organizations that have a role with respect to data indigenous data We need a collection and a list of those agents With notes from the indigenous community on them as to What exactly those agents are doing? So I'll give you an example The Australian Bureau of Statistics To our way of characterization that is a data holder. It holds indigenous data. It's a non-indigenous led organization I think that's pretty uncontroversial and And so that's in our database now certain community indigenous groups They may hold data They would have a very different kind of role with respect to data, especially their own data than the abs would So we want to characterize. We don't want to characterize We want to provide the facility For communities to characterize the roles that organizations and people play with respect to data And a note that I just took down from the last set of questions that were asked Especially for group decision making if there is a something there a piece of data that needs to Have a group make decisions about what you can do with it We need to be able to characterize that group. So we need to be able to say, you know Who's in the group what they can actually decide, etc So what you see there is a internal database tool that we're working on now That allow us to list organizations that have a role with respect to data and people And to store certain facts about them Now we're providing the framework and the tool but the the the people who actually decide what roles these organizations play With respect to particular data, that's the indigenous community So we're providing the tool and the framework the slot and then that slot needs to be filled now we can Fill in some of those slots. I'll just mention the abs But when it comes to controversial ones, for instance are such and such people the right Group to make decisions about this kind of data that kind of thing We can't directly fill that in we need communities to do that In fact, even the roles that communities could play need to be determined And that's part of the stream work one that's going on stream stream one work Now just one thing again, this is going to be very difficult to see But behind everything that we do we're making implicit data explicit So this database looks like an ordinary database. It operates like an ordinary database Secretly behind the scenes actually not at all secretly very visibly behind the scenes all the information that's in there Can be extracted in a system independent way so that if we get sick of the database we lose funding whatever All that information all the relationships and information are preserved We hope not to have a situation with the on that was talked about yesterday where if the funding ends You have a huge job trying to extract data data extraction out of this should be very easy Even if the tool goes away Another part of the work that we're doing is to do with reference spatial datasets you've seen on that Cube that I showed before we've got Spatial and methodological and other potential integration points in our project We're characterizing several important spatial datasets and there's a list of a few of them there Um, you'll see that some of these already exist things like the indigenous structures within the asgs that that already exists And we're re-representing it here and enhancing it in certain ways We're only going to do a few of these in this project one more than on that list, which you'll see in a second and um It's to show methodology. How do you characterize spatial information? In a way that can be used with the other data that our projects are working with and in a way that's indigenously sensitive That's what we're trying to do. So we're not trying to characterize all the data that's spatial data. That's useful. Just some of it the spatial data is aligned with What we think of as best practice spatial data elsewhere such as Data at the ABS and elsewhere And we can use fairly sophisticated slicing and dicing tooling to cut through data that references the spatial data like we saw yesterday with the indigenous or sorry the historical place names gazettea We can slice on spatial axes and other things And what you see there is a picture of a similar sort of data set that we're working on but with biodiversity data And in time we may be able to may be able to intersect that data and indigenous data quite effectively Because we're using similar structures for cabarets and things behind the scenes The important data set that we haven't yet characterized, but we'll get to Is the Aboriginal census police districts. This is lens work. Um, the reason it's important apart from its directly obviously importance feel is that Characterizing that data set is forcing us to do certain things which are not normally done with spatial data And i'll just give one example Fuzzy geometries a lot of data out there like postcodes and so on you've got a polygon And it's a very defined polygon and it's defined because it is defined someone can say this is where it is There's no ambiguity a lot of indigenous spatial data is not so cut and dry. Um, we have to have ambiguity and so we're working on the technical underpinnings of representing fuzzy geometries and so on again as i mentioned previously About catalog roles. This is a universal issue. There are lots of communities nothing to do with indigenous that have a requirement for fuzzy geometries Yet it turns out a lot of the standards don't actually Standardize this stuff So this project is forcing that we can't really operate without fuzzy geometries in certain circumstances We can't for instance characterize police districts where one side of the district is well known It follows a river and then the other side is you know off into the wherever that's not well characterized We we have to have a a system that can handle that to make sense of our own data and in doing that We're coming up with I guess new approaches which are useful elsewhere So in this case here, this is my approximation drawn badly in google maps But this is my approximation of southeast queens land and you can see in the the small little polygon there I'm very certain that that southeast queens land I'm pretty certain that the next bigger polygon is also southeast queens land And i'm a little bit certain that the bigger polygon is also southeast queens land. So this is My confidence in showing someone where southeast queens land is you can see at the bottom there I'm very confident that southeast queens land ends at the new south wales border So that's a very sharp bit But out towards the west i don't actually know where southeast queens land ends towards the north and the west It ends somewhere there, but I don't really know So this is the kind of thing that we hope to model and again feedback to non indigenous projects They can thank us for that later Now we do have to Because of the terms of our kind of project Calculate scores for things and the reason for this is that part of our remit is to make an assessment of how well governed data sets are How well indigenously governed data sets in existing systems are And we can only do that because there's lots of them We can't just go through individually and make an individual human assessment of that We have to do some kind of systematic assessment Which then may lead on to individual data set assessments But we have to come up with you know the basic question of all the data that's in trove What proportion of that is has anything to do with indigenous people? Sandra, would you like to say anything about trove? Thanks, Nick. Yeah, one of the inactivity 2.1 Which was you might recall was scoping of indigenous data sets. We decided to have a Look in trove and use the trove api to discover indigenous and Torres Strait Islander data sets that we Thought would hit the mark we found between two and three thousand That just eyeballing seemed to meet the mark Some of which I'm happy to report Showed exemplary data governance practices Credit to Nick T burger who's been in this game doing that parody sec was one of You know clearly really well described and pointing off to Good landing pages that explained the data and explained What it was about who is about and how it could be accessed Um Other custodians Not so not so good results. Um A lot of uh metadata pointing to other metadata and not taking you anywhere Um But it was a really worthwhile exercise in terms of um engaging with the national library I think we will need to be doing that work again um exploiting uh using Auslan codes um Iatsis have been using trove as a dissemination mechanism. So there's a lot of Data there that we can Pick up and and use to sort of characterizing in bowl can pick up things and so on so It was a really useful exercise in seeing A very spotty inconsistent Data management landscape let alone um Indigenous data management landscape. So I'll just leave it at that So it's actually from that Work that sandra did that directly led to this requirement to characterize the role which catalogs play with respect to the data in them Because as she mentioned there are a lot of stuff in trove that is actually It's actually catalog somewhere else first and then trove has recataloged it and we need more finesse there Now the three scores that we've started with To characterize data to ultimately get to indigenous data governance kind of assessments are Two of them at least are pretty well known So the fair scores and there's a green tick there because fair scores are well known There are tools that do calculate fair scores already and we're just reinventing that and we're really using that as a as a as a training and um and uh, I guess, um Comparison point for the other scores which are more important to us. We're not too fussed about fair It's nice to have maybe but as mentioned yesterday not always applicable to the kind of data We're after but nevertheless the mechanisms used to do a fair score are relevant mechanisms for other scores So a big tick on fair. It's not that hard care scores Now we have a proposed methodology to calculate a care score for something This is not something that's been done before care are a set of principles They're not usually calculated. They might be self-assessed, but they're not usually calculated from the outside as far as we know This hasn't been done. It's going to be rough. We're going to get pretty rough care scores and we look to Engagement and community over time to get better care scores, but nevertheless, we will have a go at that Now the last score there that's our most important one is the idg indigenous data governance score It's pretty imaginative title. It's not quite as sexy as care and fair and everything else But it is the direct our direct assessment of how well governed a piece of a data set Let's say is from an indigenous perspective In fact, it's not even from an indigenous perspective. It's how well governed a data set is Compared to how it should be and indigenous is part of that. So I'm going to talk about that in a second Here it is Indigenous data governance. So the first question we ask about a data set is is it indigenous at all? If it's not indigenous, it could get a very good indigenous data governance score Because it doesn't need to have indigenous involvement. It doesn't sounds fine to me If it is indigenous, then we have to ask the second question, which also seems to have the number one But if so, does it have appropriate relations to appropriate agents? And you'll remember that I talked a few minutes ago about the agents database that we have that's going to characterize the agents that we know about so organizations and people And they're essentially their ability to provide indigenous governance So that database is our set of agents and we now have to work out Whether the data we're looking at has appropriate relations to appropriate agents So firstly, is the data indigenous? Well, as Sandra mentioned from Trove, we can Find data that's declared to be indigenous by looking at things like keywords language tags Is it in a collection that's an indigenous collection? So Trove has that Lots of catalogs have that they say this is all indigenous data. They put it in there. So that's a very easy one We have to find all the different ways things could be declared, but nevertheless, that's a finite thing It's not particularly difficult. The hard one is whether we can infer data to be indigenous And to do this, this is an open-ended thing that's going to be solved In better and better ways over time, but we're not going to solve it in this project But we can look at things like the text that's used the language and so on and and work out Whether the data is indigenous or about indigenous things and this is a massive area of computer science work There's enormous amounts of data classification and so on going on around the world And we hope that our framework that we're working with here permits Increased skill in that area. So we're not trying to do automated assessment of content to work out whether it's indigenous But we certainly do want to have that slot available I put the little acronym machine learning in there. Yes, we can look at text We might look at metadata text like descriptions and abstracts But in time people can look at content and make assessments as to you know The way in which this data is or what what it's talking about how to characterize it and some of that can be indigenous assessment So taken for granted that data is indigenous in a particular case We then ask Does it have appropriate relations to appropriate agents and what we're looking at here as you can see the little picture There's a data set It's indigenous in some way Does it have this relation and the relation it could have many relations But does the relation Join this data set to a particular agent with a particular role and it's the agents and roles We've got the database for and vocabaries And here's an example rather than boring you with the vocabaries. Here's an example of the indigenous locations within the australian statistical geography standard now the Way this data was put online About a year and a half ago Didn't indicate any indigenous engagement in the data. So it did in a discursive text But it didn't in the machine readable metadata that a machine would find and be able to automatically assess So all we've done is taken the written description Of the way in which this data is managed and governed by the abs and place that into machine readable form in the mechanism I've just described now whether that is appropriate or not is not for us to judge that is for community to judge But to represent what the abs is saying about the data is our task So what the abs tells us is that and you can see there's a so-called qualified attribution thing there There's other metadata too But there's qualified attribution and it says that this data set the the rights holder is the abs You could dispute that you might think that's not appropriate, but that's what they say it is The um custodian is also the abs But the special one that's interesting here is the subject agent representative So that is a group or a body that represents Who or what this data is about and their interests So indigenous locations are obviously something of importance to indigenous people And so their representative in this case is this co-access group at the abs So that is a consultative body within the abs now again We can't or I certainly can't judge whether co-access is any good or not That's not for me to make a call on but I can record that that's what the abs have said Or that who the abs have said is their group and then people can look at that and make their own judgment as to whether that's appropriate So the the indigenous data governance score here will hinge on whether That group which we can see has a relation to the data is appropriate for this data And superficially we assume that it's perfectly appropriate And then we leave it up to people to to judge whether that is in fact the case or not But what we've done is we've represented that relation and other relations to as I say rights holder, etc So that's the core of our indigenous data governance scores It's the distance between what we think data should have Based on the facts of its indigenousness and so on versus the declared relationships that we can discover about that And so if the relationship is not declared we give it a zero score if this data had no Subject agent representative at all not even a declared one Then it would have a lesser score than if it had at least some Now a small case study Sam who sadly again can't be here He is working on a an indigenous data catalogue within the australian national university and these are the points from his project plan It's to develop a machine readable indigenous data catalogue for anu They just simply don't have a centralized or comprehensive or even very good Listing of indigenous data anywhere as far as we can tell Sam's had to work through proxies like ethics approvals and projects and all kinds of things to try and discover Where and what might be indigenous data within the university He needs to create a keeping place for that orphaned and legacy data that's within the anu And then the last two points they're building indigenous capacity within again the anu and providing indigenous governance data governance and sovereignty principles Now the two things that we're doing in this project to directly assist with that one is that Using the catalogue profile and some other tooling that was mentioned yesterday the way to bundle up data We're supplying I guess a copy or an instance of that to sam to actually create an actual technical catalogue and put things in And we've only just started on the advice side and the capacity building side That's part and parcel of the larger indigenous data network as mentioned in the previous talk But both of these things will impact on his work Oh, and I'll just mention lastly we will we do expect to have an instance of a catalogue That is at least initially only available within the anu But that all elements and the way the stuff in that catalogue is presented and described will be the same as The way we do assessments of public data so that if you get hold of an internal record for whatever reason You'll see things like the subject agent representative role and standardized metadata So even though that data is not public it should be described in a similar way to public data and to standardized data So that you don't have to go and do research to work out what the hell they're talking about Last couple of slides. So we've got plans for the future in two chunks The first set are for the next 12 months and then there's beyond the next 12 months This is partly a reflection on funding cycles and things. I think it's obvious In the next 12 months, what do we have to do? We have to Continue to operate The reference resources that we've made so we have now that we've established an agents database If we can in the next few months before june show that it has value and purpose We'll need to continue to continue to operate that we can't stop operating these kinds of things if they reference data sets That's very important. That's our number one thing things established must continue to live on And we don't the best we can to make it cheap and easy to operate that but nevertheless that has to continue We have to keep testing our tools if we start doing things like creating scores. We have to actually You know put those scores to people and get feedback and And work on that and and even at a small rate, even if we're in a kind of a caretaker mode We still have to be able to do it to some extent. We have to say yes We've heard your feedback We've recorded and we either can do something about it now or plan to in the future We do want to get organizations to adopt our catalog profile and that slide I showed before about the abs stuff That is that is adopted in the sense that Our catalog profile is being used to represent information from the abs on an abs managed resource And we want to roll that catalog profile out to other data holders And this should be a fairly straightforward job for all the catalogs I've ever worked with including geoscience australia data.gov.au These kinds of catalogs should be able to adopt this profile pretty easily We also hope to work through that deputy secretary's data group and elsewhere So that when organizations that are doing a renewed catalog Effort somewhere inevitably ask the question. How should we catalog indigenous stuff? They look to our catalog profile as the suggestion is to how to do that We will in the next 12 months work on other deployments of catalog instances And that's in particular an you but there could be others And that's because we can make a mini one and put it over there and it might be hidden But it still uses our reference resources And we want to form up the indigenous data networks advisory role We know we have this role has been asked of us, but we haven't really formed that up We haven't taken questions and provided advice in a formulaic manner Beyond the next 12 months We will eventually not just in the indigenous network But overall need to come to an arrangement to do with a digital keeping place for indigenous data I think that's obvious what that arrangement is. Is it that everything comes into A special repository that I run on my system at home? Probably not Is it a distributed thing that is Horses for courses in terms of where you might put indigenous data But there are kind of standardized ways of representing that data. That's probably more likely Does it involve a renewed arrangement with organizations due to the voice? Maybe we don't know but we do need to come up with something there and that needs to be stated It's it's in our current indigenous data network Directions and we need to Keep doing that in the next you know after 12 months We do have a an item to do digital repatriation or rematriation as it's sometimes called We do need to as Kristen and Levi mentioned find data That indigenous people do not have access to and and and either give them access to it and or return it to them So that's that's part of the indigenous data networks goals And our cataloging and so on has a role in that because if we can characterize things we can work out What is appropriate to do with it to some extent We need to develop that advisory role that we've claimed with others to other organizations We're not the only indigenous Data organization in the country and we need to work out exactly who does what we're effectively with each other And finally and importantly we need to continue Establishing continue a transfer of knowledge between data professionals such as myself and indigenous groups and that's two way and I It was asked of us before what we've got out of Paying respect to indigenous elders past and present etc. I can so I've learned a few things about data from this project like the fuzzy geometries and things but also the the Group decision-making around data access and so on. You know, these are things that I learn and so it's that way And then what I hope to return back, you know principles of data management and so on. Yeah That's it I think we're we're done on slides Thank you Len Sandra and Nicholas Are there any questions? This is a question Just in relation to I guess the spatial side of thing could you maybe hit up that southeast Queensland? Slide for me, please Because this was an interesting way to view representation of data. I thought not enough. It's a Sandra question or something but In in the example you used could this be represented if we're looking at historical data and data sets that You know, you said you were confident that that first one but less confident about the the macaille reach, right? Which very clearly is not if you come from Queensland, you know But Is there a way to represent that data in saying that okay that's strong circle there from Sunshine Coast to Gold Coast? Well, 500 sources a catalog 500 sources are recorded that say that southeast Queensland whereas Bundaberg down there's A few less, you know, and then macaille down. There's maybe two sources, but it still shows in that spatial form Is that a way is that what you're getting at there because what I'm thinking of is in terms of all of that historical data that says This is such and such country here One one anthropologist said this and you know and and a way to view that in terms of How many sources are correlating in terms of you know, where the strength and part is You know versus where it's sort of the boundaries differ a little bit. You get what I mean And then also to flipping What was a what was a modern source? What was historical source? What was in indigenously and all the rest of it? Yeah, look, that's exactly what the kind of thing is designed for Really what you see there. I mean, that's just made up, you know, my impression of where southeast Queensland is but What you're seeing is there's a theoretical what we call a feature southeast Queensland And it has a specific or multiple geometries in this case those polygons And those geometries have evidence for them. Now, there's no evidence. I just made it up But in your case you've given me evidence You said there are a certain number of historical sources that specify this one and then there's you know, etc Etc. So this feature can have a series of geometries with different levels of evidence They don't have to be numerical They could you know, they could all kinds of things that you might use to arrive at evidence But then taken together that set provides a an approximation of where this area is with higher and lower Confidences and so if you've got a lot of sources say well, we definitely know this indigenous area is here because in And instead of less so it'll it'll end up like that and where there's high Where those sources all agree you'll have high confidence or where they don't agree There'll be lower confidence and you and if when we improve the tooling that set of polygons will be a blurred kind of smooth color thing The darker the color the more strong and so on but it's all informed by different things with evidence and yeah, the one you gave is a perfect example And just following on from that Because then you'd find you'd find the section where there's an outlier And there might be one or two sources and then you can determine the veracity or the cultural veracity of that source Right, you know because you look for the outlier But then there'd be an instance of course where you find that outlier and you read through it yourself with With that applied knowledge and then you go actually that's been mistagged or that's been misrepresented And that's why it's an outlier. So You need to reflect. I know I've done that with libraries and different things. I was like, oh, this has been tagged wrong Is how do we Fix that or we're going to yeah, we're going to do that with the police districts to some extent. We're going to take Our best guess at where they are and characterize them But then we'll produce the spatial layer, you know police districts and they're these fuzzy geometries and as evidence grows about that Informing, you know, positively or negatively those things will change and we can represent I mean we can represent changes of actual geometries in actual time, you know, where a flood area was but we can also represent over time where our impressions of a geometry or other thing would be so in 1995 we thought the thing was here In 2022 we thought it was here, etc. Etc. So either one of those is possible This is that kind of time changing stuff is fairly straightforward and well known. What's not well known is that That weight of evidence for different versions of the same thing and that's what's being introduced And I should say this mechanism will be contributed back to an international data standard again, not just indigenous. Yeah I'll like just quickly add there Robert. We're also really concerned. I mean and lens case study of a variety of sources around police districts and gaps and all of that sort of thing you find in historical work We're also very concerned to easily make accessible the provenance underlying Yeah, the evidence around the evidence of these sorts of decisions and exposing that in a way You know that is that is becomes Very easy to to explore that sort of well What is the evidence behind this sort of assertion in a way that probably current gis systems don't really do and in a way Very similar to the aims and goals of our colleagues In tlc map and g happen You know, we hope to work closely with them on getting better historical data and dates and times and that So lens case work with images and this and the other and various work bringing that all together It's a really interesting case study to exercise these kinds of methods. I'll jump in there too. Um, just If you bringing that up, um Raising that that was a really good point. Um, I will point you back to the the round table that we held last year in gene as well If that's of interest and you wanted to kind of expand that out a bit more The presentation that michael erd delivered in in in the round table Is directly talking about that stuff as well So you might find that interesting to sort of point people back to there Yeah, and look, uh, I mean the gis people have got really smart tools for dealing with this sort of stuff I mean, we're talking about police districts here, but it's the same thing really applies like with language Locations as well that You can triangulate and constrain these fuzzy geometries must say, okay. Well, this is south queens. I'm sorry But we know there's a point stations of macae say So so it sort of pushes the boundary back the other way as well. I'd say you can get some probabilistic assessment I don't know. Well, I don't know. Well, there's an observation was just a thought as you were talking. Um In uh, what was it in the year 1710? Here in this part of the world What sort of data were the original inhabitants Relying on and gathering. How did they gather it? How did they record it and memorize it and distribute it? There's one observation and then come 1788 how did colonial impacts affect that data and process of from a trauma point of view had had it distort all things like that and Force removals and massacres and how did that interrupt the whole flow of data communication? And I'm sort of I'm thinking 1710 would have been through song and ceremony and dance and That's the importance of those things. So But yeah, just observations. That's all I can't really answer that But I can say something that we're doing on the technical front that will assist you that sort of thing So in the same way that we've got a fancy way of doing A feature with geometries and weighted evidence for them We can do a little bit more than that too. We can say that this feature, you know, there's The yorta yorta people's lands or something something like that We don't have to just describe it in geometric terms We can describe it discursively or in a narrative form And even if we can't go from there to a thing on a map directly We can co-record these things so we can say The the conceptual object is the homelands of a certain people and this is how it has been or is described Narratively and then we can approximate that if we want to or if we need to Spatially and we can keep those things joined up And it's part of our process of making implicit data explicit to connect bits of information that should be connected So in this case the narrative form of it of a description of a piece of land Should be tied to the thing that you see in the GIS to the point that Sandra mentioned if you click on that thing It might say well actually this is based on a narrative geometry And there's imprecision there, but here is the narrative geometry You know you make your own assessment or or whatever So we do want to keep these things close to each other and represent different forms of of knowledge Even if we can't always do a perfect mapping I'm really interested in this fuzzy geometry things that I think it's We need to go beyond fuzzy geometry to fuzzy history to fuzzy truth All together and and take that that metaphor of of of the fuzzy overlay to every everything that applies to Tab regional people particularly the stuff that white fellas have written down And I think about examples from when I was working at Uluru We were trying to develop a A booklet of stories that bus drivers could tell to tourists about different places around the bottom of Uluru And in the process of developing this booklet I found like 15 or 16 different recorded versions of these stories in various ethnographic works or travels tales or whatever and all of them were different And you know some of them majorly different And then so I went around and and recorded a whole bunch of elders in the community like 10 or 15 elders Talking about these stories and all of them were different too There is a fuzziness that's about the way people tell stories, but if you looked at the core of what You know the overlap of all of those stories. There is a core Of essential truth there that's recorded in that in that fuzzy narrative We have a very data-centric answer the general mechanism that we use for this sort of thing is to separate The conceptual identity of something from its projection So in this case here, we're looking at the conceptual entity of southeast Queensland And there's several approximations of where it actually is spatially in that case it could be that there is a Conceptual thing a story about something or other and the way that that story is actually related is different So it's projection into actual narrative is there are several issuances of that So that's the general doesn't always apply, but that's a general mechanism I I guess to the second dimension to that is that One one particular story I got four versions of it one of which was the story that's told to children One of which is the story that shared between men and women One of which was the story that only men know and one of which was the story that only old men know And all of them had the kid's story at the base of them But just more and more elements that explain different aspects of social interaction with with that story So, you know all true But all different And then there were all of the wrong versions of that story that why fellas had had also written down so You know some some were true some were not true. I mean This relates to steve's point earlier about whether the questions and things are the same The essential issues to work out whether with the conceptually this is actually the same thing at all If you can at least establish that then you can say are these variations with scope for a certain purpose as you just mentioned It's actually the same story but for children for men for old men etc So that's a scoped version of the same thing where you get into trouble is when it's actually not the same thing It's actually a different thing all together and that's that's there's no specific answer to that But I would say that the logic used to work out whether this question and that Is the same has some analog to the way we would do what you're describing there I think in that case it's quite straightforward You said they're all based on this child's story, perhaps But then they have more and more detail and knowledge and so on. So yeah, I think we can do something there It's about joining the facts up in the in a well-defined manner And that was again one of our found foundational principles in our project is to do that sort of knowledge joining John that that's what you're talking about Is a very strong cultural tradition and when boy becomes man He leaves mothers to become man, but you don't become a man in a day And you don't learn to become a man in one ceremony and that's why ceremony continues as a cultural Component of passing on information And the other thing is diversity of story. Who cares that there might be five different Interpretations of this particular totem story For example a willy wag tail or cheetah cheetah. He messes in here for death for some people. He messes in here for Integrity and other things here, but to the old girls at Uluru. They see him as which way gossip but He listening into this story, but that's deadly that there's all different interpretations And they don't have to have one universal And that's something that's got to be considered too. Yeah any other questions I just have one from online I who is this online have this is uh jelena hains May I ask if there is any intention to extend the project to south australia? I think it's a national project. It's a national project. Yeah, we may have given queensland centric examples because that's where I come from But no no totally national project and we're working mostly with large data holders generally I mean this example is about queensland, but uh mostly national data holders, which of course are national not location based Thanks, and just a comment to say thanks for the question and examples To bring it together more for the use of fuzzy boundaries to look at the police district's data, and that's from janet mcdougall at ada sterling data archive And the other thing john Sorry, they you can just say yes samurai in south australia is a partner in this project So you was on the fuzzy conversation before john and when you put your jumper on you reminded me of yesterday fuzzy was he was a man And you got into all that fuzzy was he communication just there. I'm starting right um Nicholas um, can you pass on our regards to sam that was sam just said last yesterday. Yeah, I'm sorry to hear that and Um, just to let him know that we were thinking of him and then You've been an old timer for a long time in this place. Yeah. Yeah. Yeah, and and again you can see that and you feel that So we don't know whether anybody on server said it to you, but thank you for your long-standing contribution and commitment to Advancing our interests as a people Are we ready? We're on our last Session now now a little bit of gossip lunch, you know things you do lunchtime. I had uh Three I got it with me. I had three of them little hot dog burger rolls Figured out. I didn't but bloody Helen. She had six and there was hiding them and then uh, who was into the Chocolate moose or when they got the recipe here a little bit too sweet and steve. What a how many do you are? Okay, um we're going to go into our last well formal session as a group and then There'll be a planning group that will go off and meet after this from each of the project areas, but this last conversation Is going to be with Kristen and Levi who probe Discuss talk about future requirements for indigenous research capability. Yeah Some time today would be lovely Levi I guess we should start off with, um Professor Marcia Langton sends her apologies. She can't be here today. She the last two days She's been in the referendum working group an expert panel. So doing some fairly serious business on the side Unfortunately, she only found out about a week ago But we all understand why that needs to be a priority of for all things at the moment So there's hopeful things that are happening in that space, but she sends her apologize apologies to everyone But is really looking forward to hearing back on what the discussions were and what everyone has been reporting on for each of the individual projects But also really interested to hear more about what everyone thinks In terms of how we start to think about moving forward into the future particularly in the space of indigenous research capabilities so Levi do you want to start kick off the discussion? Yeah, and I guess I'd probably just to jog everyone's imagination and you know open the discussion Probably we'll revert back to a few things that have been themes throughout the last two days One of which is looking and when we talk about the the aspiration I guess for future research endeavors in the country moving forward And you know, we'll talk about the The pipeline It's a terrible word, but um, you know, what is it over the next 10 years along with the infrastructure builder? What is it that we actually need to? Actionably do and kind of expand and work on As far as improving research capability for people throughout the throughout the country So Feel free to have any any input or any questions even just open dialogue John Okay um Chelsea Bond and I wrote a An article a few years ago called beyond the pipeline We were looking at the production of health workers Over the last 20 years and the metaphor of the pipeline in this instance is the same as in in that case It's not a pipeline. It's it's a hose and nobody's bothered to turn it on Um, it's a you know, a tiny little dribble of of people who are going through undergraduate IT courses or undergraduate public health courses that will lead into postgraduate study in in quantitative science, so If we think back 20 years what was happening then uh at an you as the master of applied epidemiology in indigenous Um statistics And we got um, maybe 10 or 15 really Top-notch graduates out of that program may be more But I know 10 or 15 of them who are working now We really need to be thinking about About that we need to dedicate some resources to targeting promising young Aboriginal undergraduates and pushing them into um Into a dedicated master's course that's going to teach them some uh some data science stuff That's all Yeah, no, thanks for that Yeah, echo and agree with that Well, I also just thinking At the launch of the new center that Levi and I and the ideana in They were talking through some of the history of there was a previous iteration of the center earlier on It's now been reinvigorated And in Anderson professor Anderson was up there talking about how In the 80s, I think he said there was something like three Indigenous students enrolled in the medical faculty at the moment. We've got a hundred So, I mean things are turning around particularly in that area, but I don't know about everyone else around the room But from our perspective when we're trying to recruit when we're trying to do all sorts of things and train people up there's really not A workforce at all or enough training going into Indigenous people in the data sciences is we're finding a significant issue So i'm assuming everyone else is having similar experiences Okay, so I get there's a bit of confusion. I think between data science and quantitative data census data figures and you were talking about, you know Cultural centers that have all these videos and audios and their sort of qualitative data I think we really need to tease that apart because they're not the same things and I think most Aboriginal people that I know wouldn't care about the quantitative data They really really care about the qualitative. They really care about Recordings and you know stuff and especially now with language revitalization It's my thing that I'm interested in going back to early records is really really important and finding those records and bringing them back and using them in language programs So It seems to me, you know, you you've got all the data science stuff That's one thing you've got cataloging of existing collections and access. So Gliathus is the issue here You know access to items that exist and then there's making safe all the stuff that's in those local cultural centers That's not safe, you know, that's I mean, yeah, we know it's that risk We know of places that have been flooded that burned down and this stuff is gone So we need, you know a flying squad or we need advice and good tools for those people So I think in this whole discussion be really good to try and keep those things Separate so it gets it can get very confusing if you're talking about, you know, this and it's actually that, you know, yeah Thanks I agreed There are a lot of areas and whether we're talking disciplinary or otherwise that There are huge gaps at the moment and you're very right I think whether it be university by university or even nationally we start have to really start thinking about how we plan for that As you said data science is one area, but it's not the only area where we really need to have More Aboriginal and Torres Strait Islander people coming up and taking charge of it We've been talking about Indigenous data governance and sovereignty And that's not going to happen without Aboriginal and Torres Strait Islander people taking the lead on it Yeah, taking lead and also being supported into that leadership as well, right like that It's not a conversation or a dialogue that can okay without one without the other Um, and I will well, you know, just to kind of keep the open discussion going We'll spare back to the idea of the concept of language language is important language matters Um language does provide reality and structure over that, right? Um The only reason why I ever became a nurse in communities because once upon a time Someone told me that I could and someone told me that I had all these really good qualities that you know is indicative of a nurse And it was three constant like repeated and reminding that you know, you manifest that and And you you see opportunities, but opportunities not always like, you know, it's not the job of, you know, self-expiration discovery Everyone has a responsibility to and so when I talk about language language historically, um Has had the unfortunate ability to kind of lock people out of stuff too, right Can't be where you can't see that was mentioned yesterday You'll say like I challenge that and say like if you don't hear that and you don't see that around you in your world Like you'll never be that you know what I mean so I guess I'd open it up to like again Uh to discussion what's everyone's thoughts about like language and we're talking about like, you know future ongoing needs around infrastructure um What's the future ongoing need around language picking up on some of the Some theme that nick mentioned as well No thoughts for anyone Do we I I guess, you know, Levi's trying to say There really needs to be a rethinking and a reshaping because the way that the language has been produced and developed over time it's through a very western colonial lens and Sometimes as other discussions have been had over the last couple of days um Some words mean many things And not necessarily good things in different contexts. So thinking through also when Essentially tasks like this is going to require everybody coming together whether we're talking about, you know disciplines You know personally medical anthropology. We've got data science. We've got archivists. We've got librarians We've got a whole variety of different disciplines. We can't even talk to each other generally particularly well So, you know, we're getting to these mixed method projects or mixed disciplinary projects And we're really not that great at talking to one another So starting to think of that overarching Way that we think through these problems. I think is going to be one of the core issues that we're going to have to tackle in coming years Hi, I just have a comment from online from Dr. Jelena Haynes Who says that she's noticed the big gaps in South Australia, especially in library and information science Which is an important field of field to encourage indigenous students to gain knowledge and skills to record and preserve their knowledge in their own context I think we can all Agree that we all experience similar things like that. Um look LAM sector is not my sector Perhaps we've got someone in the room who'd like to comment on that No, potentially. No. Is there anyone online? Or perhaps we'll move on from that The um, I was I was going to sort of talking about that, you know Um, I've always been as as much as I might be Um, big note in the other day there Well, about saying, you know, I've got a couple of degrees and that anyhow I I've always felt like I'm a liaison officer or an interpreter And I think a lot of indigenous People in all no matter what level you're at That you're still nearly like an A liaison officer or an interpreter and And yeah, because I'll tell you I have two boys at home And one's an engineer Um at Boeing and the other one's A sound designer and I don't know if you know anything about sound design They're the ones that sit there and put the sound into movies computer games a whole lot and We sit at the table and have dinner without telling you too much personal stuff And and we don't actually say much to each other Because they're sort of like, you know, he's The engineer is the engineer Another one's a sound designer and they're really into the computers and They don't say much And I'm and I'm actually a social worker So rave on a bit And so we got I and and that's what we're talking about here a little bit And I'm saying we're all Aboriginal people You know like at home there and um, and then I had my wife in the mix and she's Irish British background. So We have an interesting Conversation and sometimes I feel like that that's what's going on here Is we got all these different backgrounds And we're trying to work on indigenous issues And then coming from all these different disciplines And like in the house feel I I I think I'm located there as a social worker, but Love, you know, I love my work of a mob yarning up So my point about this is saying that We're trying to bring a whole bunch of different people together to To work this out And and hopefully we can because but we can't leave out The liaison officers Yes, and and let me say there's also our Remember someone had a shirt there um Where they say I'm I'm safe or something poor indigenous people. What do they say? You know Ally that's the word. You know, I'm a safe ally, you know and you can't I think That's also what we need as well. We need safe allies with us That we know that we're going to trust them with our stuff They're not going to do us in So if if we can find people like that to work on these projects, I think that that's a Good thing. That's a bit of yarn and I'm I'm doing I'm doing the old yarning to tell a yarn Perfect example Yeah Just following on from that point about the safe allies is I don't think Nowadays you can be successful if you're not working in the best interests Of mobs and communities. However On the flip side of that from a monitoring perspective Perhaps there's a conversation around big institutions need to apply some monitoring mechanisms to Assess and find out who those people aren't aren't safe allies anymore or who slipped through the radar Back in the day and are still highly acclaimed Academics who don't hold the best interests of indigenous people but have made a career You know a career off Using indigenous data and those people still exist. They are still very much alive And for communities, I have heard communities explicitly say I'm just waiting for that I'll bugger to die so I can actually go and do some proper work So just on that flip side And I think that's a really familiar tale as well The amount of times that you go into different communities and there's always a story of a pretty dodgy researcher coming in and not doing so great things in Whatever context it is and look, I mean A lot of it Probably ignorance. We can chalk it up its history We do need to know about it and if people are still operating in that way today Which we know that there are there's every good reason for people to be a little bit suspicious And to actually ask the researchers to do the things that we really should be doing Yeah, I thought that last point it's interesting and I I do think there's some other changes in the in the broader Kind of research environment that that go to the point that you made to robin One of the things that was happening when we were looking to have an era and we won't we won't be having an era assessment So so for people who don't know what that is That's a an evaluation of the quality of university research And there was going to be one in um 2023 And and one of the things that it would have had to incorporate is that we now have um new research codes And and so so research is evaluated against research classification One of the one of the classification standards that we've been talking about And and and there is a there is a new indigenous. It's called indigenous studies And and there was a proposal to actually think about how do you evaluate research in indigenous studies? And and in particular, what do you need to do to consider issues such as community benefit community leadership community engagement Now we're not um We are not sort of going to have an era and we're going to move probably to a different kind of process But some of the signs were there that around something like that which kind of shapes stuff quite importantly for universities You know, there are some changes in other parts of of this system that that sort of also go to some of the things that we're talking about Yeah, look, I think there needs to be general institutional and national change when we think about the way that we do indigenous research It's very very clear That there are a multi a multitude of elements that are required when you aren't working with indigenous people We're all working with indigenous people when we're talking about data Um, that isn't necessarily the case in other fields of research or other areas For example, you know, if you get your standard ARC Grant You do not have a budget there for the extensive consultations You do not have an extended timeline for the extensive consultations that are required You do not have the requirements in there to have the onward going conversations and feeding back into the community The research that you've done and I think, you know across Australian research, we need to have a really significant think about that some of the philanthropic Grant bodies are starting to factor that in but very slowly Which means that we are going to have rubbish research that continues on the disrespects the people they're working with That doesn't actually factor in anything to do with indigenous data governance or indigenous data sovereignty because it doesn't allow that And it's very very disrespectful. And I think as a nation we really need to stop take that on board and Significantly advocate for change, you know, whether ARC You know the NHMRC wherever it happens to be We're doing research in these areas, you know, researchers are the ones who are going to have to advocate for that change It's not going to happen by itself I had a comment from Dr. Jelena Haynes who said that she Agrees with Dr. Stephen That we need to include the voices of liaison and integrating storytellers voices and they're lived in experience in the discussion Yeah, I'd probably Yeah, kind of, you know feeding back on the dialogue and responding Yes, we need to include those voices but that liaison stuff in the gap work there that is a gap And it's not also just the job of like the liaison or like, you know, the indigenous person in that position Because Italian that's burdensome as well, right? So this is the reason why we're asking questions What's actually missing? What are some suggestions? And it's not oh, let's load bear that up Of the of the person in that position that ain't it It's just like this is the reason why, you know, we're asking questions It's just like from, you know, the amount of knowledge and from knowledge expert holders and stuff in the room It's just like just one really pragmatic thinking It's just like I don't want to be this exhausted by the time I'm like, you know, Marcia's age It's just like I don't want to get there and have the same question and trying to be fine the same solution Because if that's the case like And you know, when we talk about research and stuff, I think it needs to actually precognite before we get to like the level of a master's I think, you know Data, you know, we're in the fourth industrial revolution data is in in every part of every job So it's just like no, no, there's actually a focus that needs to shift This is my opinion anyway We need to be doing that like at the very core essence It's just like how do you interact with data in your in your world? What role does that play? You know what I mean and then like then the specialization and that stuff So I'd like to hear everyone else's thoughts about that I have just got from my limited experience A little bit of a comment around the resourcing and the time and you're The point to not wanting to burden people with the responsibility around education Extension liaison, etc Of my experience of the time and energy it takes to come to a shared understanding between the researcher and the to's of what indigenous data is What indigenous data governs governance and sovereignty is and I think that there's a role just to support that around education of both researchers and to so we have a shared understanding beyond indigenous data as Traditional knowledge Like a broader understanding of indigenous data and your rights and interests I think would be there's I see roles for that. There's a real need because our people out there doing it out of the kindness of their heart and their spare time or really extending themselves and burning themselves out so That comments lovely. Yeah, thank you Any other burning thoughts in the room that you want to talk about? Um, I just want to come back to the concept of the ally Um, and we've been talking about pipelines over the last couple of days How do we start to think through a pipeline for actually teaching people how to Be a good ally Instead of just thinking maybe they're doing the right thing, you know, actually some concrete pathway forwards for that Um, I know that you know, I know that iatsis has the core cultural training, but it's kind of an opt-in kind of situation Um, yeah, what can we do to further ally ship? I guess Yeah, and you know bouncing off that comment to then Jenny I guess like we need to probably reevaluate the actual very, you know, very definition of ally ship You know, because it actually does imply that you do renounce and actually align yourself with like, you know So That's not necessarily a case. I don't think we you know, I don't think there's Like across the board There's a full spectrum of people that have that level of like all that humility and stuff and practice and you know renouncing that and actually Wanting to seek solutions and actually wanting to be informed And wanting to work in a way that's like, you know, non-injurious. So Just people as well And I guess taking it in another direction I was just what's that? Oh, there's another question Oh, sorry Sandra. Thanks. Thank you. Um, just before the round table I'll ask you. I forget which journal it was. I think it was an international Journal of regional health and the editorial board of that journal adopted a policy position that said we're not going to publish any research articles unless the researcher can demonstrate That care principles have been followed Um by and put the onus on the researcher to show that and In a past life when I used to be a trade union official dealing with issues of technological change Consultation gets you so far, but let's face it. It's not easy to do this stuff It does take time to be done properly And that is not something that either universities or funding bodies want to hear And they need to be being advised on Researches should be able to submit in their applications sufficient Resourcing and time to make this happen because perhaps the consequence of It not happening is that research doesn't get published universities will suffer reputational damage Some things that they really we know they really do care about Um, so that's just the old trade union official coming out in me there a bit that that you know, um Possibly may be trying to exert influence on commercial publishers to adopt these kinds of standards might win a few hearts and minds Thanks Sandra. I think that's a great idea because we know that there are a lot of publishers out there who are also asking for advice and how they can engage with them Indigenous data governance and and protocols and standards. So that's actually a really helpful Piece of advice or the only thing I worry about is the researchers who are trying to publish in there and then they're not getting the funding But in turn it does push back on the universities just thinking about the squeeze in universities Yeah, true We need this systemic change, you know, absolutely Okay, so I was going to take it in a slightly different direction So we've got in the room a range of different researchers. We've got people from community controlled organizations or have worked in those backgrounds So I assume we've also got data custodians in the room When you're working with indigenous data, what are the things that Often come to mind that would be a barrier for you to ensure that you are either connecting with the right data custodians or Ensuring that you are following particular protocols that you think should be in place What what do you think are some of the barriers that we need to work on? So I I work on collections in Australia that come from Papua New Guinea, the Pacific and other places And these are collections that would be lost if they weren't digitized and the people in those places can't get them. So When you're doing this stuff, there's always risk And I think the problem is that we often allow the risk to outweigh the benefit And so for what we're doing the benefit is that this stuff Can get back to those people that there's a catalog online They can find we take it back to local cultural centers and agencies for them to find these materials Which are their grandparents, you know, I mean it's that kind of thing But the risk is, you know, we're the colonialists. We've got this stuff. We're in Australia They're in the pacific So, you know, we you have to balance all that stuff up But if you don't balance it up and you just say, oh, the risk is too great. We're not going to do it Let's just forget about it. Well, they're the ones in the end that will not get access to that material. So I think, you know Being an ally, I suppose isn't just everything that You know, everything that's non-colonial is great. Everything's colonial is bad. You've got to weigh these things up So I think it's complex and you should be prepared to engage in complexity as well Yeah, thanks, Nick. Absolutely. I think we all see that in our work and it does take courage. I think sometimes to sort of absolutely say, well The risk is not going to be as high as what it is if you don't do the whatever action it is to whether return the data or material There's always going to be a way that Our institutions will often ensure that they place the legal risk frameworks around us that makes it really problematic, but I think some of those things we need to think through a little bit better as well particularly when we know that a lot of indigenous knowledges are being lost And look, I think it's the job of all of us to ensure that We do as much as we can to preserve and maintain the knowledge that still remains Yeah, and I'd probably say like, you know, you're looking at diversifying and maturing your risk appetite in that sense then, you know what I mean No, and I mean that wholeheartedly Just like, you know, a lot of Aboriginal communities and stuff. It's just like we're we're seen as like, you know High risk taking whatever but that is by default the position that we find ourselves in right It's just like I'll give an example of the COVID like pandemic. I think that's a very good thing that'll illustrate it Is that we had The aqua sector the Aboriginal community control health sector working over above and beyond Getting stuff done working through it We did not have the time and the ability to all the luxury I should say To wait for a clinical governance model to come down from department So in in that sense, we use data. We used research skills in an echo To diversify and we built that clinical model three months ahead of government Do you know what I mean? We we have to head towards risk. We move through risk not around it So it's just like, you know, I'm going to say, oh, maybe we need to like diversify mature our risk appetite That's kind of what it is So being brave, you know, so Any other comments Particularly to do I guess with the type of work that we're in. Well, I know not everyone's working specifically Indigenous data, but indigenous data is usually in most data Um I know that a lot of the people that we've worked with before are very very hesitant to do anything about it because they're very unsure So they may want to be allies, but they don't know what to do about it And are terrified to kind of take that next step. I really want some sort of assurance Often I think I mentioned it before the IDN gets all sorts of queries ranging from really basic to broader but The IDN in particular we do not have the resources There does not seem to be any sort of national body to deal with any of this I mean, we've got a team on the ground of what Levi maybe Altogether Maximum but that's in kind. I mean in terms of funded through the idea and I think we've got three three positions So we definitely don't have the capacity there doesn't seem to be resources going in anywhere else But there is an absolute need in order for allies to actually become allies For some of this support to be going on so On this your risk risk management is part of It's a fundamental part within a standard governance Framework, but what you identified Levi is things that are not being identified. It's the cultural risks and what defines cultural risk versus standard risk And that's the level of cultural We need to precede words like psychology and sociology with cultural And what are the cultural risk elements that fit within a shared governance framework in this? framework that you're working on For the and call it a shared governance is a reciprocal thing because you can't achieve what you need to achieve I was talking to Robert Yeah, the day and we drew on a napkin If you look if you put the Aboriginal people that you want to engage with in a circle in the middle of a page That's where you want to get to But what you've got to understand is that to get into that circle There's a whole level of layers of complex trauma If you go into a stolen generations Circle the trauma is even more complex than what it would be to be bought up in Cherbourg or Warabinda Versus Christian mission versus government reserve. That's the level of detail We've got to get to truly understand how to engage authentically and with integrity and honor in this space so When you come in from western governance, you're talking on an outer circle where it's about performance and accountability starting with leadership and strategy policy procedure risk management Um monitoring reporting. That's all your your standard western Stuff that's on the outer circle, but you still got to get into that inner circle now the people like Levi Who can go into that middle circle and understand the trauma and live the trauma day in day in you Whether you like it or not you you you still got it when you're working down the road in correct Okay, so that's the old notion of 24 seven You can go home, but he he can't Because it's clicking over and to you it'll be still clicking over to but different dimensions of it So to get out of this process there's a There's a need to build trust mutual understanding respect and trust My time working in Aboriginal affairs. I can say that You know, I've heard the terminology that it's an industry. It's an industry And a lot of people do well in this industry It's like in the incarceration space. We're the highest incarcerated people in our own country. We're impoverished And oppressed in our own land. We're welfare bound on our own country So We are an industry and then in the incarcerated space people get paid good money to work in incarceration to to Keep us under control Historically you'd be controlled in a mission or a reserve, but now they can't do that. So the next alternative is prison It's clear logic to me that the system is operating the way the system wants it to be there is reason why we're oppressed And the opposite to Privilege is oppressed So when you start to evolve forward, you've got to really peel back the layers And really get to the core because you're trying to get into that in a circle But there's a series of steps That'll allow you to come halfway in and you'll never ever expected to get into that in a circle of trauma, but he's in there He'll come out and meet you halfway, which is what happens But you'll go back to a western Government system university system framework of thinking he'll go back into the community So somewhere you've got to understand him and he's got to understand you but you two are allies You have to value and respect the contributions that you each bring but you've got to agree to meet Somewhere in the middle Which is not in his space and he will never get to your small he could get to your space and you could get to his space But That's the big picture. Yeah, and that's why I have a point about shared governance 42 years Same shit different day In many ways and that's not saying this of here Um The amount of government rhetoric I can take you back to a report. I remember Steve you remember that in the miller report It was all about making a concerted effort As so it's concerted attempt to change the social economic circumstances of our Brisbane Torres Strait people It's the same shit different day So again, what you're on a halfway you're on a journey But you have to get the shared governance framework intimately Unpacked and restacked and you've got all the key players in this group If you catch a one on picture that all come from your risk analysis stuff Journey we might ask Yeah, yeah I have a another comment from dr. Jillina Haynes who apologizes because she can't be on the video right now Saying that she works alongside the elders for 20 years and our concern now is to access financial support To digitize recorded knowledge that has been recorded and some of the elders are no longer around I think that's also a common thing in in many communities. I know that where we work It certainly is happening everywhere the Lack of resources in that space is really quite shameful. I think at a national level Because as as we all know Indigenous knowledge a lot of it is all there is is on these tapes And once they're gone, what is the deadlines 2025? Um, if it's not digitized by then it's gone forever So I think you know, that's another thing where there really needs national advocacy to ensure that there are appropriate resources going into digitization But once we digitize it There also needs to be the infrastructure to both score it and ensure that it is Held in perpetuity and looked after and managed properly in a way that is right for the community who belongs to it I guess one of the things I'll probably mention is ultimately the audience is wanting to function and you know really kind of enrich Community of practice So yeah, we would encourage anyone to kind of come along to that as well Like I said, only one of three staff that function the IDN And it's like I'm a big boy, but there's so much of me to go around being spread Spread and possibly fit at the moment And obviously it'd be way for ignorant of me to think that I had all the answers and all the solutions That is clearly just not the case So what I like, you know, one of the lovely things is is day in day out. They do work besides Kristin Um, so all things feel really possible Only 10 jaded the other part of me is like starioid optimistic. So Yeah, I would love to like fold other people into that space as well Um, and because the thing is it's it's about collective knowledge to write for a shared goal and But benefits average and trust are on the people collectively benefits everyone, right? It's a positive curve theory, so Anyone with final thoughts before we get it wrapped up? Yeah, just oh I know I seem to rave on a bit. It's just a social work thing but um When you you made me think of um It's it's one of like I I lecture lecture to in In bringing people together and hopefully win a few allies In uh, social work and also got some psychology students This some this semester, I think It's about this thing of um having Uh, when you know like about that allyship How do you like and we're always hoping that we get people into that space? We're at the end of the course That they're on board as an ally and sometimes One one of the things we say about people is that you're not losing your culture If you learn about Aboriginal or Torres Strait Islander people What do you do and is you're actually gaining? And that's the same thing with all different cultures You're actually expanding your world And you're gonna you're actually gonna know about other people's world and you'll be a better person I know that sounds really hairy fairy done it But but when you actually take on other people's cultures And you start learning about that Suddenly your world expands Rather than putting the wall up going Oh, how come these Aboriginal people what's this referendum thing about? What do they want a treaty for you know like People put that wall up. You don't need to put that wall up You need to expand it and say hey come on. Let's go for it See what happens here and see what I might learn from Aboriginal Torres Strait Islander people But then also what you may share with them From your culture I have you know, I have friends from overseas and met a lot of really good people from the Ukraine And you know like you think of think of that You know and they're And and sad to say I've been invited over there a couple of times, but I haven't gone and And I'm just saying to you expand Expand just you know expand yourself How you kind of work with Aboriginal Torres Strait Islander people I think that was a brilliant way to wrap it up unless we oh, no, we do have one more Um, this is slightly more general comment I'd like to make but it does clearly relate with this discussion Particularly Levi the discussion about risk I think anybody who's been in this room for the last two days Can have no doubt that all the strands of this enterprise we're joined in are deeply serious About the risk management aspect of handling data And And this is a topic I've actually discussed with Steve mckacken along the way and I think I mentioned it to Jenny sometime I think We should consider the possibility that we can renegotiate With universities Some aspects of who takes responsibility and what I'm thinking of here is that people who are Committed to putting data through our processes Should get benefits in the ethics process from that the university should be able to say You're putting your data with ID and fine. We take every box because we know that it will be managed responsibly That's a really good point That was the way to finish Thank you guys sorry if I got ranty Yeah, thank you everyone um Levi I'm very proud of where you've got to and what you've achieved Coming from where you've come from Now that you need to hear those things and the same was young robert the same it's Back is the same. It's about Just acknowledging The work that you're doing and giving you the authority to continue to do what you've got to do Not that you need our authority, but you're just being respected for all the effort that you do That's what our older generation need to do with the younger ones build each other up and not pull each other down Yeah, so I'm from here um I'm going to just give you a general I pulled out words. I've been writing down words um Over the last couple of days To try and piece it all together for me The first one is holy shit. No, it's not. Um It was starting to be talked about there um In the data space, this is just me asking questions Who has been privileged and what was the opposite of privilege in the context of data And the opposite of privilege as I touched on just then was oppression um when it comes to allies It's important thing. One of the things that ally ships all about Is that you've got to understand your own privilege first Before you can actually go out and engage with others. You've got to take the first step You'll make mistakes But you've got to learn from your mistakes. So in the data research commons and indigenous research You're going to make some mistakes. But as you said, Simon These are all here together. You're all allies already. You just got to consolidate and strengthen that um I think there's a need to Define and sell what value data in a modern society Brings to first nations communities across our diverse landscape Because a lot of Aboriginal and Torres Strait Islander communities on the ground Would not know understand half of the things that you're talking at at the levels that you're at And you need your liaison type persons that can go halfway and meet So they're the they're your internal allies So sell the sell what value it and the value that it can bring for a whole range of other related things for the community shared governance Obviously needs to be trauma-informed healing focused and culturalized for whatever better word Um and it needs to have its own sort of a reconciliation focus Up to this point Aboriginal peoples access to data Has been probably limited But you're actually creating an opportunity to enhance that your capacity to access all the necessary data that you need Has been limited. It's the evolution of how technology works um Three projects operating and a few other components that we've heard over the last two days and I wrote down to myself um Respect for collaborative ethical relationships have to exist between all three And you need to be streamlined You need to do what you do Value your own autonomy project autonomy, but you need to recognize that you're running parallel with each other to a common a common end Mutual trust You need to have established mutual trust across your projects with each other You need to think about notions of And you haven't seen this yet or this language, but kinship leadership or kinship led project approaches By that son, that's going what you were saying is is to valuing the the unique and special skills of every individual here What you bring to this process for the benefit of all You have to support each other if one of the project teams are suffering Well, then you the others help out in some way you do that in your own autonomous groups But in the context of the three projects that we've talked about here You've got to keep an eye out for each other in my mind and I might not be right, but So there's mutual trust Respect for collaborative ethical research That's accessible streamlined trauma informed healing focused And underpinned or guided by your fair and care principles That's about all my waffle What I want to do before I head you back to Jenny Jenny and Michael Now I haven't heard Michael. He's been as quiet as a mouse all day yesterday, but I do want to hear him by the end but not just yet I want to go table by table and just give me You take away thoughts for the last two days And I want you to hand it to my deadly sister over there the blonde hair because you had something to say earlier And it was worthwhile saying and we're going to go Be honest Hi everybody, my name is Alicia Um, and I'm also a social worker. So I apologize in advance by waffle. Um over the last two days, I guess I've heard a lot of things that have been out of my depth or beyond some knowledge technical things I've learned about um Methodologies for doing things and lots of awesome tools, but um the thing that I kept coming back to is I I wonder if the real depth of if the people in this room realize there's people at the end of these Methodologies, so I work for a community controlled organization Um, I've spent a long time working Um with vulnerable families on the front line And some of the things that you're doing have significant real world impact to people life changing things like reunifying children Kind mapping families when people thought their identities were lost. And so I'm just Yeah, this last two days has been brilliant for me around um Seeing all these big systems big picture things but just trying to remember that there's communities at the bottom of this community controlled organizations On the front line of gathering some of your data, uh, we're the ones that are gathering that um and You know, we we're the ones that go out and hold community trust And so I just I think about that and I think about How we can apply some of these tools and make sure we're meaning community. We're maintaining community trust. It's important yeah And that's fair. That's Feedback on the run Any other comments from that table? Look at this. Oh, john. You're going. I knew you was going to say something john No But you need a strong cultural foundation and an understanding appreciation of What ancient ways can be applied of modern days in many ways and To me, that's that's the conversation needs to start That you start with to actually truly get an authentic shared governance framework and approach Now this lucky gentleman over here is sitting by himself Take away message for the last two days. Um Sometimes I reflect I've got 11 year old and we live in the same house But we like live in different worlds, you know, his his view of the world is so different from mine I wonder if we have the same kind of thing between Indigenous and non-indigenous. We live in the same land, but we live in different worlds And we maybe we need to build a bridge across those worlds Well, who's going to start that's the conversation, isn't it? How do we build the bridge? Yeah Yeah In this room that are from different sectors that are like White fellow black fella as well. We this is the bridge in the commons and the research commons is the beginning of the bridge of the two worlds I was really inspired by what was said yesterday around like there's only black fellow ways and there's Space for you in it and like the invitation For all of us to participate in the one world, which is australia. I'm not sure that it's two different worlds. Yeah Ancient ways modern days. It's as part of the answers in building a strong sense of kinship and kinship respect back to humanity But again, it's it's you're right here. You've got a whole group of different people who are all Individuals that are all invested in this space But you're all family so your family in this data commons space But you all have your own Individual idiosyncratic ways and methods of thinking and being and that's okay. That's what makes it interesting but support each other and build the bridge and It's the old saying that you know, we can say that that's okay, but we're Aboriginal people See it. It's like the liberal party ideology was that everybody starts on the same line. You're familiar with that The only one that starts on the same line were white males land hunting males predominantly so Black people are back here migrant Indian people Back here women back here So we've got to break down those barriers and get better understanding that sort of stuff But do it together We are one we are many from all the lands know if we come that would be a good start Yeah, you're over this way right I was just can I just can I jump on that that bridge building thinking I was thinking if you're going to build a bridge there between those two worlds. I mean one world has this A whole heap of knowledge in there in terms of sustainability. I mean it's sustained for 80 000 continuous living years so there's a lot of reason to build a bridge and access that knowledge in many respectful ways And I'm sure from that world that if we're living in black fellas, we'd build that bridge, but you see on our land Our access to our resources are all locked up We can't how do we build that bridge in that sense? when the purse strings are held by governments and successive governments who've made decisions that have effectively Caused the problem that we are in now when we talk about indigenous data governance And that more broadly and the problems we face with that today It is the imposition of colonization as to why we're working in this space If that didn't exist, we would know very clearly because those governance Lines that had existed for thousands of years would have continued The sharing of knowledge of indigenous data would not have been imposed upon too. So I think the impacts of colonization is pretty important Is a keystone to the and we're all aware of this because we're all working in that space and and it does impact You know what we're what we're doing, but I just wanted to raise that point specifically Just around access to do and the very one of one of the very simple solutions comes back to Putting black felt the right black fellas in those positions to make those decisions as well You know to work with other black fellas, too So it's you know, they talk about indigenous lead indigenous first those kind of concepts putting that at the forefront Putting them in in control of those resources to build their half of the bridge To use the analogy Well, value value the contributions. I'll come back to this table over here and we're heading this way There's a challenge for this table over here work out who you want to speak Um, I think I probably just want to say that um old is new and new is old. Um, I've actually got a lot out of today in terms of context and scene setting really and contextualization and a lot of this is um, it's new but it's old and it's all intertwined and having a look at it in terms of government policy and processes Engagement with stakeholders building in Aboriginal community control because that is so important, but also building in um, our communities building in our indigenous worldviews and building in our families in terms of some of the things that you've been saying um, I'm Here um with iatsis and I think in terms of some of our engagement Some of our engagement has been old But it needs to be new and we need to really kind of embed ourselves in some of this process and this discussion Going forward in the way that we probably should have been um, historically over the last little while So yeah, old is new new is old. Thanks. Yep, and that reminds you again of the ally ship where In an ally to be a good ally you you're allowed to make a mistake You can you will make mistakes. There's no no such thing is perfect And no one has all the solutions But you have to work together, but if you continue to make the same mistake Well, then you've got a problem Uh, and I'll go back to my gentleman print at the back And I should have said from a cultural kinship point of view. You're 11 year old son is my 11 year old grandson in a in a kinship context Uh, if I was your your brother Well, he's my uncle father to your son if something happens to you I've got a responsibility to steer your son On the right path. That's what kinship's about and that's what this process can be all about So I'm a new lucky duck you get the own Hey, I think I'm gonna sound like Pollyanna here But I'll go ahead anyway a few minutes ago I said that if you've been in this room for the last two days, you couldn't doubt that we were serious about certain issues And I like to expand that and say you couldn't be doubt that we are serious about tackling these incredible problems and I think on the basis of having a group of people who are serious about the problems and have goodwill in their hearts We can make progress and we are starting that journey Yeah, well, you actually well and truly on that journey Now Simon's been very quiet and I want to hear his lovely soft gentle spirit and voice come to the fore take away message Simon before I hand you back to Hey You know Simon, I'm looking at you Michael and I'm talking I'm talking looking at you and refer. Oh shut up, Brad yeah I don't feel like I'm the right person to speak because I'm actually very much a beginner on this journey and I did want to just Convey my respects to the people in the room have been working such a long time um I think probably what's required here And um, you know, if anyone from Encriss is listening um, I think what we're starting in this room is something which will be useful for the whole landscape of research infrastructure, not just the haspit And I think what You've been really conveying very well Is it's it's when you talk about a bridge? The the traffic goes both ways, right? um, and there's a lot of talk, um, I feel about um communities needing to be helped That's one direction But I think the other direction is there's a lot of help that can come from communities for this country um, and it's There's a lot of wisdom And there's a lot of good practice that this country would benefit from enormously And if you sort of it that way Gosh, wouldn't you invest in the other side of the bridge? Yeah, right. Yeah, so I think Seeing it as a two-way bridge is really really important to to value that so You know a great experience And again, it's it's valuing the individual knowledge skill set that you bring each each of you bring to this whole thing I've got a set skill set you got a set skill set, but I need you as you need me sort of thing um Same thing this and I don't think we put enough emphasis on How we value other people's knowledge and skill sets? That's that's important part of this process And Neil I'm going to come to you as the esteemed old timer On this table Len Len, yeah I'm hitting that time of the day. I'm calling charlie browner Yeah, thanks look This was there's been a really strong indigenous focus on our discussions and that's really great I suppose if we step back a bit We're really talking about building institutions which are addressing the data needs of First Nations people and the Australian community as a whole and I think Just to tick it off. I think we'd need to Particularly pay attention to the fact that the program of work that's the jetties sort of lettuce on Has some notable gaps and I keep saying the one that worries me most is that There's not enough focus on the humanities and there's almost no focus on the arts So Next stage somehow that's got to be addressed in terms of the indigenous program What I keep coming back to is how all this is going to fit into the voice and truth telling especially I think The role of information and data is going to be absolutely central in that and I think the sort of institutional arrangements that support the voice You know, I think well Today we're probably looking at the seeds of what that is those institutional arrangements are going to be so, you know Let's build on that. Yep And before we leave that table, I'm going to go to that handsome looking fox there that's scrubbed up like across between Scott Morrison and Bob Ork You ever said nothing while you've been here. I want you to tell us what your take Yeah, look for me, I come from an organization That looks at or talk the conversations around infrastructure and data In a totally different context. So this has been really rich for me to sort of come back to that understanding that That is not just about like ones and zeros just from the technical point of view that there's embodied in that a people a stories a culture So I think that'll be something interesting for us to go back to our organization Because it's very IT focused Uh to to sort of again come back to you know, why are we why are we supporting as an organization research What are we trying to move towards and and help reinforce and not lose sight of that and that's I suppose a better future I think I'll leave it at that It did well Now I can go to this Young lady over here who's also sat down that same spot and young lady is kind. Thank you My name is Sarah poking horn. I'm based at RMIT, but I've actually just recently arrived from Canada So my takeaway is just All the learning that that you've made possible. I certainly came to the right place Just in terms of getting oriented to this whole space and the conversation here And of course coming from Canada, that's kind of where my mind is also Going because of course the conversation is not identical But a lot of the a lot of the challenges are similar But a lot of the opportunities and strengths and maybe potential connections are Parallel and and similar as well. So so just just thanks and and it's been a real pleasure To sit back and take lots of notes and try to absorb and you've been joking about the acronyms. But yeah, that's that's a That's a real That's a real thing, but uh to no one's surprised. So just just appreciation on my part Yeah, lucky duck again I'm a lucky duck loner on a table I would just like to say thank you for opening this up. Yeah, open invitation. It's been a real honor and um Lovely for you to say that um, you know as allies were allowed to make mistakes and that's a real relief. Thank you because I think it takes um Some you know guts and gruel to kind of walk together and um Healed together um trauma moving forward and putting You guys, you know, you know Position of power because I think you've got a real opportunity at the moment. I think there's a real Taste for you know at e-research Australasia this year or last year It felt like every second talk was oh and we want to embed the care principles. So I guess um, you guys indigenous lead can shape how you want that to look and I look forward to Yeah, joining your journey. So thanks Whatever comes up or being partnership with yourselves Uh, this second Foxy look hasn't said much Hello So i'm actually part of the team uh in the indigenous studies unit Um And these are the types of conversations that we have every day Almost every day right about data and i'm a data analyst. I'm a sociologist quantitative sociologist So we talk about these issues every day and it's really good to see That these conversations are now kind of being had in a in a group setting with people from different backgrounds um, I wasn't here yesterday. I was um, I was busy with um Uh, basically doing the same thing but with uh with other people in the office, but just coming in here today around noon Seeing people who are part of our team so len and sandra and and nick john all the way back With whom we're also partnering as well as some old colleagues from other departments around the university who are here It's just a great It's great to see people involved with data From different backgrounds and again kind of it goes back to what christen was saying That data has different meanings to different people. I know as a even though i'm into kind of the social science space The way I think of data analysis is at times very different to the way That somebody else um in a maybe an adjacent field uh would would think about um data So it's just a very good feeling. It's a good feeling of camaraderie And also being able to have these conversations outside of our Small office over there on gratin street. It's uh very good to be here So thank you very much and uh, thank you. Um, and thanks to christen as well. Who's my boss, but also, um, someone who had a lot to do with um Organizing this session as well as janny and her team as well. So they can't before you hand that microphone All right, can you say this On the cold and rich Chicago born and another little baby child was born in the ghetto Now do you reckon Simon come on And his mama cry Believe I wasn't that your boss is sitting there. You're going to do it now We laugh with nor that rotate. What's your takeaway? Thanks so much. Yeah, look, uh, yeah, just really happy to be here I think it was a really great opportunity for me To learn much more about the other two projects of DACA and the indigenous research capability I you know, we we just talked today about about kind of building bridges and connections and I really like your your words around, you know, helping Each other out and so on and I do think, you know, for me It's I've only been involved in in the iris project for, you know, about 12 months and obviously I think for the other projects as well as for us the focus has been on You know, we haven't had much time. So, you know, we just try to demonstrate something, you know build something between, you know The time that we within the time that we had basically until the the end of June this year You know, we we all kind of worked within our kind of projects to to to do something and to be able to demonstrate something and, you know, as you Could have seen within the iris project. There is enough complexity, you know, navigating the six working packages and so on But I do think now is the time to actually kind of maybe start thinking and move a little bit beyond the projects And, you know, as we as we as we move into the future, you know, there's this really, you know, I can see possibilities, you know, for example, you know, the Indigenous research capability, you know, obviously stuff around data governance That we'll have to engage with the geospatial aspect. So, you know, it's great to see those Presentations and and, you know, find out about those things and and these are the kind of things I think connections and how we can work together and and, you know, obviously Everybody's going to have to maintain that stuff internal focus as well, you know, you keep keep pushing things forward But then, you know, how how how can we reach beyond the individual projects and try to do something together? Michael and I had a sort of a discussion around future As we're talking about data and research data commons now in 2023 But what was what will the conversation be in? 2073 to think ahead the past present future And that's where I think your focus should be also And this is this way this this um event here the symposium is a great, you know, great thing to be in this Yeah We'll go to the back bench and I'll go to the middle gentlemen sitting in the middle Hello everyone I'm Jonathan Bollen and I I teach theatre and performance studies In a school of arts and media. So it's really interesting to hear the conversations about about the scope for arts and creative arts data to Come into the commons and I can certainly see the potential for that. I also think that That there are Approaches in the creative arts that are that really resonant with some of what I've been hearing and learning about from First Nations peoples and indigenous perspectives on data about the way data is Not only, you know countered and visualized but also embodied and shared and and communicated through song dance and And storytelling and in gathering as well in bringing people together. So all of those The social architecture dimensions is something that's shared with the With the creative arts and I I think there are also A lot of scope to be working with Creative people to make sense of data. That's that's one of the things that I've been working on I'm looking at the the theater companies who are who are who are using data to create performance so rather than writing scripts or or Or or you know gathering stories. They're actually turning data into into creative works of art. I think it's a really interesting time. So Thank you for the opportunity to listen into these conversations Thank you very much to each and all and Joe Joe Thank you online. Who's online online? We have a few comments from online. So Dr. Jelena Haynes says ideas learned from the from the two days were respect trust mutual relationship and reciprocity Ancient and modern days needs to coexist and work together Maria Weaver said as a librarian working Relatively recently started in the data common space I've gained a lot of learnings from this symposium and I hope to continue learning and expanding my understanding More over time. I appreciate all the comments and insights from everyone. Thank you so much for the sharing and openness Jennifer rumble dr. Jennifer rumble from She's a Camilla Roy woman from at Newcastle University says that she really appreciates this space to hear groups talking and sharing respectfully hearing too that people are looking at lost and orphaned data By bringing together innocent and bringing together in a central space For many purposes keen to hear more as the IIRC and IDN great work. Thank you And thank you to uncle Grant for your wisdom in guiding us And uh, Robin Burgess from ARDC Is the research data specialist in data governance at the IDC and he's thrilled to hear About the importance of data governance being mentioned so much And he looks forward to hearing more in this area and learning more and contributing to future discussions Thank you. Thank you. And can you think uh, Just the online audience. It's always difficult not to be able to see but thank you for their consistent contributions and questions that they raised which was very important Um, thank you Joe For the work that you do behind the scenes to get this stuff. Um, Mary Thank you all so because um, you do a lot of things behind the scenes I wouldn't recommend eating six hot dog hamburgers at lunch anymore. No, she didn't But you're part of the team that helps Jenny and and their things in a kinship based Arrangement It's easy to get up here and do what I gotta do But I rely on you and that's an important thing that needs to be valued So there are things that we need to embrace across this whole process What's your name sister? Yeah, you know, I'm looking at you. What's your name? Sonja. Good on you. Sonja. You've done a deadly job too And our brother fox up here. Look, what's your name? Steve Good on you, mate. Thank you Um, thank you each and every one of you for your participation over the last two days It's been always it's a pleasure to come and spend time in the presence of unique and special souls And it's my pleasure to and that's not a pleasure It's just great for me personally to see people come together and treat each other like family And respectful and that's what it's all about. That's what caring sharing and respect Is all about and that's who we are all children of the universe here Some of us have had a bit better than others But we're all in the same We're all in the same ballpark trying to do the same thing and that note I'm going to hand it back to janny and michael Who's going to close it up? Peter. Peter's going to come up to It's all yours, jen I Did want to touch on what grant said about kinship when we started on this path Back in june 2021 Um, I think you know, we were quite a diverse group of people We still are a very diverse group of people But what we have come together with is kinship And we are very blessed that we all work well together And we've got one another's backs and I think that that's really important. So, you know, we're a good team And we're willing to take more people into our family of has an indigenous rdc So thanks everyone for coming along. Thank you to everyone who helped out There's more food out there We're going to have a little planning session just for the partners So if um, but thank you to the rest of you who've come along And hopefully we'll see you again next year and thanks to everyone online as well Thanks to my ardc colleagues for their assistance. So yes, thank you everyone