 Thanks Sarah. Hi everyone. I wanted to just continue on from a statement that was made in the leader's panel about what the rest of the world could learn from New Zealand and I said it in an earlier presentation today. I'm really humbled when I come to New Zealand. The engagement of white New Zealand with its indigenous residents, inhabitants is just something I think certainly Australia should take note of. And it's quite humbling to see the genuine deep and committed engagement you make with Māori and islander people. Thanks Sarah. As Sarah said, we released our data set. The collection. So we've had a collection since 1946. It composes 157,000 items. They are predominantly single channel moving image, predominantly 16 mil and VHS. We operated for most of those 70 years as a lending library, educating schools, universities and film societies, not only about film but through film, other subjects through film. So we have an incredibly diverse and eclectic collection. As part of making our collection more accessible, we released our data set as Creative Common Zero license. We released it as both an aggregated tab separated file and also as JSON. And in this first release it was 40,000 records. They comprised the 40,000 title records of the old lending collection. Underneath that sat multiple carries in many instances. Following the release, we walked across the road really and met with RMIT, one of our tertiary institutions in Melbourne, and invited the School of Design Master's students to use our data in their data visualization. This is just a little bit, you're not meant to read these particularly, but I just wanted you to see how sort of raw and comparatively uninteresting the data looks in these forms. And hopefully then you'll be astounded at what the students did with them. So here's sort of the raw data, here's some early conceptual thinking. So as I say, we approached the RMIT Royal Melbourne Institute of Technology School of Media and Communication and we provided a scaffold set of tools to help kickstart the students' thinking. One of the important things we did during the collaboration is we also attended the students' presentation sessions to offer advice and present feedback. That was a really interesting process because one of the things it gave the ACME staff was a real sense of achievement. They, as many of you will know, working in archives and we work in a windowless basement, can be pretty unfulfilling and quite repetitive and it feels like you're never going to reach the end of your task. Here were my staff, the collections staff, walking into a tertiary institution where the data was being celebrated and it was being celebrated in whimsical and playful ways, but also in really insightful and thoughtful ways. So it had this fantastic feedback loop for the students as well, for the staff as well. So as you can see the student, oh we're one slide behind. Right, so there we go, we did that. So this is so flash here, having on these multiple screens. The student benefit was that they could read and interpret the data and draw narratives and communicate complex information in accessible ways to non-specialist audiences. It was real world data and they were real world problems that we were trying to overcome in terms of opening up the accessibility and that was really valuable for the for the lecturers that were running the course. One of the other things we did is we took them to the vaults, so we wanted to connect the data with the physical nature of the collection and there's one really effective visualization we'll get to, but I thought it was really powerful that they understood that this data, particularly in this you know digital savvy world, this data actually linked back to tens of thousands of physical things. The benefit for ACME was that it helped expose the richness of the collection, but it also, so it made the collection more discoverable, but internally it also gave us new marketing ideas and it also illustrated as we'll get to later some of the acquisition gaps, so it presented some opportunities for us to fill gaps and make the collection better. So what happened? We had 56 infographic student projects and as I say they range from the whimsical to the political. As I'm sure many of you know with data visualization they revealed patterns and correlations obscured in text-based analysis and as I said they ceded new marketing approaches for ACME and they identified gaps in the collection to target for future acquisitions. So what did they look like? Some of them were just a bit of fun. So this is monster films and the percentage of the entire collection that deal with monster films. This is how many films on vampires, aliens and other monsters we have and this is what they look like distributed over time. So it's just a bit of playful way to engage people with the collection. This one looked at three key phrases, somewhere, nowhere, somebody, nobody, nothing and something. What it revealed is in fact it's more important to be nobody at ACME than somebody. This I think really illustrates a couple of things, the power of the site visit. So this is a digital representation of our compactus. He's stripped away the shelves. He's then taken a very small slice of the collection, in this case documentaries, refined that further down to documentaries dealing with music. Now it's only 1.4 percent of the collection. It's an incredibly small component of the entire collection, but then he's illustrated the richness that exists even within that less than 2 percent of the collection, which I think is a really powerful representation of what the collection has and it's led to those conversations about how diverse, eclectic and far reaching our collection is. The one that really resonated with our female CEO was the appalling gender representation in our collection. Those figures speak for themselves and that generated an immediate response from the executive team to begin to address that imbalance. This just seemed like a nice marketing campaign, this one. I'm not sure it's entirely successful in its iteration, but what the student has attempted to do is look at films with an astronomical subject, space race, planetary issues, telescopes and so on and try and represent them as a constellation, if you will. They were design students, which was really interesting for me. I'm a film archivist. I work and I move in film archivy circles. I don't have a lot to do with design students. This student based their design work around Japanese iconography and then use that to represent the number of films of different languages that speak to the subject of Japan. Fascinatingly, we have one film in Polish that's on Japan. The nature of this exercise, I guess, was to present the collection in a visually appealing way, but it also revealed to us as managers of the collection that we had these really curious little things in the collection. I haven't looked at it yet, I must admit. There's what it might look like at a bus stop, at a train station, out in the real world. Part of their brief was to try and design not only the visualisation but how the visualisation might be applied in real world settings. This has been really great for our marketing and communications department, and particularly for us as the collection is migrating from that of a lending collection to that of an archive. With the recognition that we have many rare and significant items in our collection, it's been really useful for the marketing team to see how they might apply these in the real world. We have a really good collection of Australian experimental films. This particular project illustrated that we are actually missing a number of key works that would make that collection exceptional, so that's given us an opportunity now to go out and target them for acquisition. So you can see some of the applications within the institution as well. This one I really like, this one's I think the kernel of a really interesting research project. So this student looked at films made on military, agricultural and environmental subjects and mapped them to the Liberal and Labor Federal Government periods. He in fact looked at two particular periods. They were the Hawke and Howard years which were relatively similar periods. What he revealed in fact was that during the Labor Party's federal reign more films on the military had been made, more films on agriculture had been made by the Liberal Party and more films on the environment had been made by the Liberal Party. So there's some interesting stuff being said there I think. It doesn't tell you the nature of the content of those films and I think it warrants further investigation as to whether for instance those films were anti-military in nature or pro-military in nature. So there it is graphically represented. So it's a quick talk isn't it? What did we get out of it? We wanted to make public knowledge more public and more easily consumable by the public. And as Brad Halock one of the RMIT professors that oversaw the project says there, an awareness of such repositories of public knowledge in all their diversity is surely an important force in political hygiene. As the headline of a 1960 newspaper article about the then State Film Centre now ACME reads, this collection was at the time already keeping democracy alert and well informed through film. Furthermore he goes on to say, while admittedly not unique in any of these capacities data visualisations of these types are powerful examples of design practices through which we might care for others by seeking to reveal a world less veiled and in all its diversity. The thing I wanted to say to you is it was really easy. As a cultural institution we just let our data go and I think in my 30 years in this industry there has certainly been a tendency by institutions to not want to release their data till it's perfect. Well guess what? It's never perfect. It's never done. The work we do is never over. But we just let the data go and we got all of this reward back for that relatively easy effort of just setting it free. Not only did we get the reward the staff experienced by having 56 students enthusiastically engaged with our collection, we also found gaps that we can pursue for future acquisitions. It sparked ideas in our marketing and communications department as to how we might advertise the richness of our collection. It got us thinking about our collection in entirely different ways. It was really important for us to have this collaboration and I would encourage those of you from cultural institutions to make that collaborative effort with your tertiary institutions. They're just around the corner. They're always looking for opportunities to exploit data like this. This is a relationship that we will now carry through. We've already released all our cinema screening data for these students to do visualisations on. We operate two cinemas at ACME. So we've got 15 years of screening data, both attendances in cinema, film shown, times of day. All of that data is now being used in the next term for visualisation projects. As I said, it kick-started research and acquisition opportunities for us. And of course the more we as cultural institutions release this data, the more it can be aggregated together and we can start to explore the relationships between our film collection, our national library, our state library, our artworks, our photographic collections and so on. So I would encourage you to do it. That's all I've got to say. Happy to take questions. Thank you very much. Are there any questions from the audience at all? Hi Nick. Thanks for that. I have a reaction to your presentation and then two questions, if I may. I thought I was really blown away by what the students came up with and I think it really speaks to the allowing that design perspective and sort of like the counterpart to some of the things that Huck and I were saying. When I think of data visualisation as a manager of a website, I immediately think of kind of programming and how you solve kind of programmatic problems in taking data and visualising it. But this is kind of taking the other angle of it. And I just wanted to say that some of those examples that students came up with are just really prompting all sorts of thoughts in my mind about how we tell the stories of our collections. I just have two quick questions. One is what plans have you got to keep the data sets up to date and any thoughts on the sort of format in which you do that? And secondly, has the exercise prompted any changes in your description practice that might yield some different avenues of these visualisations? Yeah, good questions. So yes and yes. They're the short answers. I agree with you. It was really interesting to put this data in the hands of designers first, you know, not archivists or librarians. And I think, as I said, I thought that was very rewarding for the staff as much as anything. We do plan to release the rest of our data. We use Vernon, big shout out to Vernon, good Kiwi product. Our data is slightly complicated in that the traditional lending library is structured around a title carrier relationship, as you would expect in a traditional library. We have Mad Max as a descriptive title record. And then we have all the subordinate carriers as related records. And then we have time based media art, objects, games, all sorts of other things as object records with parts. I was talking to Paul Rowe about this yesterday. In many ways, that's a bit of a historical hangover from the lending library days. It does make amalgamating the data a bit more complicated, which is why for this iteration we just released the title records for those films, those entries that have title carrier relationships. So we are working on the next release. And something that's really important about that, what is missing from that data set at present is the work that ACME has done in the community making digital stories, collecting personal memories. One of the reasons that's really important is we have a very strong collection of utilitarian films from the 50s during that massive period of migration, European migration. And similarly in the 70s, what ACME has done in the last 15 years is collect stories of those migrant communities. So there's a real opportunity now to connect historical data with more contemporary data and begin to examine how that migrant experience has changed over time, for instance. And in answer to your second question, our cataloging activities have changed really dramatically in the last five or six years. Unlike many traditional archives, which I feel that the catalog record, the catalog information has been largely around stock management. It hasn't been big on descriptive data, it's been big on components of a title. ACME has been a lending library for most of its life. Our data set was all about the description. We wanted people to know what were on these films because we wanted people to borrow and look at these films. What we're discovering now, particularly as we move into things like collecting and digitising home movies, is that we need to go beyond just the descriptive paragraph or two and virtually into shot listing. We settle on sequence listing rather than shot listing. So our data is now much more exhaustive. We're practicing trialling machine learning for some of that to try and take the incredible onerous task away from humans as much as we can. So they can do the interpretive work, but the machine can do the identification of scenes. That will open up a whole new set of opportunities for the visualisation. So we see this very much as the beginning. I think this relationship with RMIT will stay in place. We are coming to the end of our first year. There's every intention that will go on next year for both terms and I would think we'll just continue to build on that relationship and the data will get better. While I've got this opportunity and the cameras pointed on me, I'd like to thank Brad and Jigga and Andy Serrong without which this project would not have been possible. I was just curious how you made the initial connection with RMIT and sort of yeah well you answered that you're going to carry it on, but are you looking to open the data for other students or other research group to be able to use your same collection data? So the data's out there. It's on GitHub. It's zero commons licensed. Anyone can download it. We've been in recent discussions with another tertiary institution and their school of design to potentially do the same sort of work. How did we decide to do it? I can't take the credit for that. Seb Chan, formerly of the Cooper Hewitt, I'm sure many of you will know him, he made the first initial approach to RMIT and what was very gratifying is he recognised the quality of the data that was in our collection and that you know data sets like this are out there. Cultural institutions are releasing their data. What is special about ours is it was all about accessibility right from the beginning. That's what we started doing in 1946. The entire rationale for the data was built around accessibility. So every film is mapped to the Library of Congress subject classifications, 26,000 of them. So the data was very rich. So that was what sparked his interest. He approached them. They couldn't wait to get at it. Oh, no, I'm good. Hi. You talked a little bit about how you had to provide some or you provided some tools for the students to start with to get them started. I wonder if you can tell us a bit more about what you actually provided and set them up with and also the tools they used in terms of interpreting the data. Sorry, what was the last bit of the question, the forms? The tools that they then used as part of their project in terms of interpreting the data and getting to those visualisations? We sort of gave them a bit of a roadmap into subject classifications. So it was just really planting the seeds of ideas, making a few initial connections between the data. So for instance, the one that showed time, almost there, this one. So we suggested to students that they could look at a subject and that the subjects were actually provided in a separate CSV file and that they could then make connections with films on that subject through time. So it was just to spark some ideas and I will admit I wasn't actually aware initially that we had provided that little bit of encouragement to the students. What came out of that is that some students really didn't do much more thinking than that. So we got 56 projects. They were all interesting. Some hadn't really gone much past taking a subject and mapping it over time and then perhaps mapping it by format. So that was a surprise to me. I thought, oh, look at all these students. They've all done this thing. Isn't this great? And I will, in fact, the technician that released the data had sort of provided that for the students. Which was fine. I think it was useful. It can be quite a challenge to look at a blank piece of paper and try and come up with all that thinking. So we really just provided a scaffold of ideas around which they could begin their exploration. So it wasn't much more than that. Most of the students just chose to work with the CSV file rather than the JSON, which again was fine. We might be done. Sarah, what do you think? So thank you. It's been great to attend NDF. This is my first time. I've really enjoyed it. Thank you for having me. Thank you for listening to me. I've got to go to the airport now. Join me in thanking Nick once.