 Welcome everybody. It's Jerry Ryder speaking from Anne. I'm sitting in for Karen Visser today who's unfortunately unwell, but it's my pleasure to welcome you all to the webinar today, and particularly to welcome Doug, Rahul and Dave who are going to be presenting with us this morning. So today's session is, as the title suggests, in our data citation series. It's the second in our data citation series, and we have four in the series altogether. This one is going to be a really interesting session because it's some real case studies from the Australian National University and the Australian Antarctic Division about their journeys, if you like, in implementing DOIs and data citation at their institutions. We'll start off with a presentation from Doug Monker, who's the repository manager at ANU. Then he'll hand over to his colleague Rahul, who's a systems developer at ANU, to talk about their experiences. Then we'll hand over to Dave Connell down in Hobart at the Australian Antarctic Division to talk about the experiences at AAD, and then we can come back to some question time after that. Okay, well, hello everybody. My name's Dave Connell. I'm working at the Australian Antarctic Division. I'm in a section of the AAD called the Australian Antarctic Data Centre, and our job is basically to look after all of the data that we collect from our scientific program. So, here's our little case study. I thought I should start off with a bit of background information about the AAD and the Antarctics, to give you a bit of an idea of what we're dealing with. Basically, science has been very prominent in the Antarctic over the last century. Props are a bit more so than exploration, as what people tend to think of most, I guess, when they first think of the Antarctic. One thing that's probably a little known is that Douglas Morton was actually offered a spot on Scott's trip to the South Pole first, but he turned it down in favour of running a scientific exhibition instead, and that sort of pretty much kickstarted the whole Australian sort of science process down there. There's quite a lot of international sort of science that goes on as well, and that all sort of culminated in signing the Antarctic Treaty in 1961, which was originally by 50 nations, and it's still running to date. It's one of the longest running international treaties going around. And one of the tenets of that was that all scientific data needed to be made free and open to anybody who wanted it, and that's sort of one of the real sort of foundation stones, I guess, of our data centre. It sort of gives us our mandate for going and getting everybody's data and pushing it out to the world. Some other things about Antarctic are obviously a long, long way away. It takes about a week to get there in a ship from Hobart or about 6,000 to commercial jet, so it takes a little bit of your time to get there and back. That tends to make it very expensive. It can cost anywhere between many thousands of dollars to send somebody down for a few weeks or hundreds of thousands of dollars to send them down for a year. So if you go into all that sort of trouble, then you want to make sure you look after your data appropriately. Here at AAD, we have a multi-themed science program with many hundreds of scientists involved who not just base it here, but they're also at universities across the country around the world and also for other state and federal government organisations. Over this last year, we had 60 projects running for the summer, and all of those projects will generate data, which has to be collected, managed and published by our little section. So basically, we do a lot of different science, and there's an awful lot of it to keep track of, and then we've got to make it all available to the world whenever everybody wants it by download or via other distribution services, if possible, and it'll cost a lot of money, so we don't really want to have to waste any of our data by having to go collect it again. So that sort of brings us around to DOIs and where we're going with that. We made the decision last year that we needed to start creating DOIs for our datasets, and obviously if you're going to create DOIs for datasets in Australia, then you need to talk to ants, so we approached them for assistance and they pointed us in the right direction, shows which paperwork we had to fill out and what sort of things were required in order for us to start creating them. We did briefly think about creating DOIs for our metadata records as well, but decided against that in the end as it was perhaps a little bit of overkill and wasn't really going down the right sort of road. We figured just having them for datasets was enough, even though the metadata standards that we use actually have a spot where we could put a DOI for our metadata records into. We did anticipate back in 2012 that there would be a strong level of interest from all our scientists in getting DOIs for their datasets, so we figured this was certainly going to be the right thing to do and we were trying to do the right thing by our scientists as well and also provide them with service that we thought was going to be amazing. And then finally, after getting around through all the jumps, all the hoops and hurdles and all that, we made our first DOI earlier this year at the end of January. Okay, so there we go. So our little rules and guidelines that we have, we don't automatically create DOIs for all that data sets, which is something that I know that CSRO do. Well, I think that's what CSRO do, I could be creating if I'm wrong. And many other people are probably possibly thinking about doing that. We decided against that and decided to put the onus back onto the scientists to request a DOI for their dataset. So if they wanted one, then they would sort of come to us. The only exception we made to that was that if, for example, a user was out there who had found a dataset and said, I want to use this data for my work, I want to correctly cite it, does it have a DOI? And in that case, we would then sort of go back to the scientists and say, look, somebody wants to use your data, we'd like to make a DOI for it. And generally, that sort of thing is so far as where it's happened once and it worked out quite acceptably. We also, one of our other business rules is that we only create DOIs for stable datasets, which is one of the reasons why we don't automatically create DOIs for everything. We wanted to make sure that the dataset that we were creating a DOI for was finished and it was all done. There weren't likely to be any more updates to it. Also, that data had to be publicly available. We didn't want to, because quite a chunk of our data is still sort of locked away. Some of it's because the scientists are still working on it. They haven't finished publishing yet. So we want to, obviously we don't release, publicly release data that they're still using. And many of our other datasets are commercial and confidence things like, things like phishing data and the like. Also, the last little rule that we initially decided on was that there had to be a one-to-one relationship between our metadata records and our datasets. This is sort of because the way we've sort of gone with this, even though the DOIs for the dataset were actually attaching it to the metadata record, because the metadata record describes the dataset and that was sort of the easiest way for us to organize it. As a result, sort of some of our metadata records actually linked to several datasets. This is just how the scientists like to organize their things. But in the case of DOIs, we decided that because of the way, the fact that the DOIs actually going to be attached to the metadata record, we had to have only one dataset in each metadata record. And so far, that hasn't been a problem that's been working out, okay. Okay, so getting our little message out there. So first of all, back in December of last year, once we had become an official sort of part of dataset or of ANS, I guess, so that we could create DOIs, we put a little press release out around our organization, say that we could now create DOIs for everybody. And then we sort of sat back and waited for the influx of DOI requests. And then we waited a bit more, and then we waited a bit more again. And eventually we had one single request, which came out in the end of January, 2013. So we decided that this little approach which is waiting for the scientists to come to us was perhaps not as useful as we thought it would have been. So what we decided to do was to beat the bushes a little bit. So when scientists were submitting data to us, because we now have that sort of fairly well entrenched with them, they understand now that when they correct their datasets, they then hand it over. So when they did that, and if they indicated that the data could be made public, then we would ask them, would you like a DOI for your dataset? And generally once we'd actually explained it, they sort of went, oh, that's a great idea. I would love to have a DOI for my dataset. The other thing that we found that this also did was that when scientists were finishing up their projects and now handing their data into us, we would ask them if they wanted a DOI, and at the same time we could sort of say, by the way, if you want a DOI, you have to make your data public. Something that's often been a little bit slow to happen, but because they were sort of suddenly sort of dazzled with the magical properties of a dataset DOI, they agreed to make their data public sort of that much sooner. So that worked out quite well for us. One of the things that I guess we're hoping for with dataset DOIs is that eventually they'll obtain sort of like a critical mass in our organization, much like metadata records did, I guess. I've been here at the IAD for about 15 and a half years now working on metadata for most of that time, probably about 14 of those years. And when I first started doing metadata here, it was a really hard slog. People weren't interested in it, they didn't want to do it, they viewed it as an unnecessary overhead and it was just a waste of their time. But eventually as we had to keep persisting and the metadata collection grew and it got to a certain point that suddenly it started becoming very useful and after that my job got a little easier and people started handing their metadata in and their data sets in to do that without being prompted. And we're hoping that something like that will happen with the DOIs. So once they start seeing the dataset DOIs out there and around and in papers and in publications and on websites and stuff that they will start coming to us more rather than us having to go to them. One of the other things that I guess we're also trying to sort of entice them with is the ability to track their data sets. Something we're not really set up with yet but we're hoping that will become sort of the more sort of common in the future. So that can then see who else has been using their data and where else has gone. Okay, so this is still also part of our little message. One of the things we thought we should do is create a little DI logo. We haven't done this yet but this is a little screenshot of what sort of thing that we were thinking of. This is a little snapshot of our homepage. I'm not sure if you can see my mouse on the screen or not but I'm just going to wave around in case you didn't get rid of that. But down here in one section of our website we have this little latest edition section. One side is always devoted to maps and one side is always devoted to data. And what we were thinking of doing is that every time one of these data things appears that has a data set DOI we stick a little DOI logo on the top of the corner so that suddenly it gets a bit more exposure on our homepage and people are seeing it a lot more. And actually up here on the top you can see the little press release on the 13th of December 2012. That's what we sent out. We're noticing we've got DOIs. Come and get them. The other parts we do as far as our DOI message is we're trying to get in with a bit of early education. So for all the projects that our scientists apply to do they have to create a data management plan. So if this is a very new thing we've only just started at this last season and all projects in the future will be using these. And part of the data management plan is this little bit on the screen you can see here. It's just a little bit of education. It sort of lets them know that we can create DOIs for their data sets once they're finished. And that's sort of correctly citing their data. And we have a little check box in there that asks them if they would like to be able to correctly cite their data once they're finished. And we have that set to yes by default. Hang on, I'm just going to cough for a minute. Okay, sorry about that. And at the moment this is just really, like I said, this is just education. So it just isn't tying them into anything yet because they're writing these plans so they can click on any data. So we're not expecting to see any data from this for the next couple of years. But at which point we can at least go back and say, well, you said you wanted a DOI and let's get on with it. Okay, so just a few quick screenshots of some of the tools that we've put in place. So at the moment anybody can request a DOI for one of our data sets but only I can approve them at the time. So what happens is that somebody would come to our website, preferably the scientist in question who wants to get a DOI for their own data set and they will search for their search for their data set and then they'll arrive on this little page which will ask them to formally request a DOI. And we provide a little bit of information here as to what happens after they do that and also how their little citation will look. So what happens then is that the request comes to me and I've got a couple more screenshots here. The one at the top, you can see is part of my little admin page and I can see on the left what the data set is who's requested it, which case is probably requested by me, it's just a W1 and also a date and the time. And you might notice that there's a slight bug with the dates which we haven't quite sorted out yet as I'm pretty sure I didn't request that in 1970. Then on the right I can either delete the request if I decided it's not a legitimate request or somebody's just testing the system and if it's okay then I'll submit the request off to And. One thing I do before I submit that request to And so is I have to go check the metadata record to make sure that it has all of the appropriate fields in the data set required. Usually there's one missing from our records which hasn't traditionally been entered and that was the data set release date so usually I tend to go and enter that in before I submit the request. If I submit the request and it doesn't have all of the required fields then I'll get a little error message back telling me what's missing and please fix it. So then once I submit that request off you get the I get the DOI back instantly which is very nice and then I can go insert that into the metadata record and then load it into our database. Down the bottom of the screen there you can see a few datasets for which we have DOIs but you can see the DOIs listed out there and also I have the other tools there for me how to deactivate these DOIs or update them if the metadata records or datasets slightly change. Okay so obviously if you're going to have a DOI you need to have a landing page this is just a section of one of ours so you can see basically at the top the first thing we threw in was the citation so the site is going to initially quickly see this is how I cite that dataset there's a link to the dataset on the right and then we've done on the left and down the page we start moving through some of the basic metadata elements there are a few more that are off the bottom of the screen that you can't see and also we provide a link to the full metadata record which is on our catalog okay so basically where we stand now so at the moment after having been available for DOIs I can see for a few months we currently have 17 of them in our system a few of those have been user requested but most of them have come about because I've approached the scientists and said look if you finish this off now would you like a DOI for your dataset um there is a strong level of interest in DOIs but this most seems to be mostly once they know about them so at the moment many scientists seem to be blissfully unaware of them but once they actually find out that we can create them and then what they're for then they're generally very excited about it because I think anything that involves citing their data they tend to quite enjoy so we do expect to get quite a lot more of these in the future so as I just said the biggest problem is a lack of awareness from our scientists but they do know about DOIs for publications so they are at least aware of what a DOI is so it's not a foreign concept to them and luckily for us we're not the only drivers of getting DOIs for datasets as many journals are now asking scientists to submit their data as well as as a paper so that often sort of becomes the requirement of getting a DOI too or it's something that scientists would like to have a DOI for their dataset before they submit it to their journals okay so that's basically sort of where we're at