 Okay, so we'll just have a look at some of the footage that is actually part of the program. What we're looking at is a propaganda film by the National Socialist Movement of the Netherlands, which sympathized with the Nazis after they invaded the Netherlands in 1914. And it's showing this wonderful, bucolic life, boys playing around, and the text actually says, the voiceover says that these will be the leaders of the future. Well, we all know that didn't happen, thank God. And so it all looks very nice, this is typical propaganda. It's actually part of the collection of the Netherlands Institute of Sound and Vision. But as I said earlier, even though I'm from the Netherlands Institute of Sound and Vision, and you probably wonder, why am I presenting war collections? Well, I'll tell you about this. Can we go to the slides please? We've provided a lovely old world yellow background for you, just to remind you that we're sort of the only presentation from Europe. And the, so what am I doing here for the network war collections? Basically, I'll tell you a little personal anecdote, because I do love facts, but what I even love more are the sources that back up these facts. And this brings me to my first slide, which, because I woke up on November 10th, 2016, and I just couldn't believe that I had woken up in a post-truth society, where facts just didn't really matter. Someone could have said something the next day completely deny, ignore it, deny the next. Science had just become another opinion. It wasn't about facts. Even people coined terms like alternative facts, which in Europe, I mean, I'm sure this audience would agree with me, was really, I mean, we were all baffled. So that was actually the morning as well that I had to go to my Amir conference in Pittsburgh, which I don't think I attend any of the talks because I was basically depressed in my room pretty much all the time, because I had really had an existential crisis. I was wondering, what the hell am I doing in an archive? Why would I want to be an archivist? Why would I do this? Because facts don't matter anymore. Sources don't matter. People just ignore them, and you can get away with it. But I snapped it, I snapped out of it, I'm here, and I actually chose to become an advocate of facts and reliable sources. So this has really been a really important driver for me to do what I'm doing today and to keep on doing what I'm doing today. So when I was asked in 2016 to become, to chair the board of an amazing program project called Network War Collections, I immediately said yes, because basically what it does, it brings together various sources of archives, collection museums and so on to tell new stories about the Second World War. And it's not just new stories, but it's more relevant, more reliable stories, stories that are actually based on facts, on different sources that are knotted together in a very clever way, something that Lizzie will tell you, will tell something more about. So as I said, Network War Collections is not what I do for a living. It's not my day job. My day job is the deputy director of the Netherlands Institute of Sound and Vision, which is a 50-50 publicly funded institution in the Netherlands with around 300 employees. But I always say that Sound and Vision is more glam than the other glams, the galleries, libraries, archives and museums, because we are actually an archive and a museum at the same time. We have about 1.5 million hours of audiovisual assets ranging from public broadcasters to commercial broadcasters, amateur film, animation, documentaries, government films. But we also collect YouTube, social media, Dutch games and so on, basically to reflect on the media history of the Netherlands. And what we do, we have very strict service level agreements with broadcasters, media professionals from the Netherlands and abroad to basically get the content out there to deliver it for programming, for news items and so on. And at the same time, but we also push our content to educational platforms. We make it available for academic research, but also for journalists who actually still bother to actually check facts. So that's what we do. And as you can see, the building is quite spectacular. It's just like 30 kilometers outside Amsterdam. And we also attract around 300,000 visitors a year that actually come to Amsterdam using, which basically reflects our collection. As early as 2008, we actually decided to embrace the two-speed IT ID, which now seems a bit like obvious, but at the time it wasn't so obvious for an archive. Because we realized that for an archive to have impact on a digital age, we needed to reinvent ourselves, but not move away from what an archive should do. So what we have done, we have a very good data infrastructure. We have 30 petabytes of material that is stored in a trustworthy digital repository with a data seal of approval. We have an off-the-shelf MAM system to comply with all those service-level agreements that we have with media professionals for the throughput of all the material, the ingest of all the material, and so on. That's 5-1, which is also used by BBC, Al Jazeera, the media asset management system. But on top of that, and I think that's really what is relevant to the network work collections, is we also try to be very agile, not so much in terms of the methodology, because we're sort of agile-like people, we're not too strict on that. But what we mean by that is that we've built sort of a flexible back end, which acts as a layer between the MAM and different portals applications that we have. Because we believe that with that flexibility, that gives us enough flexibility to do more as an archive than what a traditional archive would actually do. And I would actually dare to argue that these things that you can do extra with that flexible back end might actually be more important than the ones that a traditional archive does. But I won't tip you. And for the agile development, we basically use three very simple principles. Smart, connected, open. Smart basically means automate whatever you can. I mean, it's not a man versus machine thing, but machine just do it faster. Of course, there are errors, but you know, the amount of data they can actually, and that's been touched upon by some of the other speakers, the amount of material that they can actually process is enormous. It's something that we cannot compete with anymore. Connected, don't work in isolation, always make sure your infrastructure, whatever you develop, allows you to be in touch with other organizations, other institutions, other collections as well. And even from a collection metadata point of view, try to connect other people, collaborate, because there will always be more knowledge about your own collection outside your organization than you can actually organize within. It's a very arrogant thing of museums, collection galleries, archives to actually think that they have all the knowledge about their collection. That's really not true. And if there's anyone who says differently, I would love to talk to you during the lunch. And then we have the third principle, which is open, which is kind of twofold. First of all, we encourage you to use open-source software in this agile development. Of course, we still have a solid foundation of the man, but with the agile development, we actually try to use open-source software wherever we can. For example, we try to be the launching customers of a lot of academic research spin-offs that have developed, for example, new facial recognitions, speech, text. And technologies. And another one is not just open-source software, but also open up your collections, share book categories, be open. And this is exactly why we teamed up with the network board collections as sound and vision. Because suddenly our collections were reached, we reached users, we reached other user groups that we would never have reached. They knew things about our collection that we didn't know, and they actually did things with it that we would never have imagined. So that's a very important thing. And what you see is that, for me, the two most important principles in this context of network board collections are connected and open because they're basically the foundation of linked open data. And linked open data is also the foundation of one of the most important principles of the network board collection program and the way we work. You know, the goal of linking open data is basically to make an extension to add a data commons to the web by publishing various open data sets as RDF triples. But also by setting links, RDF links between different data items in these different data sets. So over the years, what you see here is one picture shows 2014, number of every dot is like a big data set, and then that's 2017 right next to it. So you see the number of RDF triples, but even a plurality of all the links, they just are growing exponentially and I think they will continue to do so. I think opening up data and content has a wider political and economic context. Linked open data is basically a main driver for innovation, for knowledge, but also government transparency. Luckily in Europe, the EU really encourages or even obliges to a certain extent public organizations that produce data, information, content, even audio-visual content, to publish it as open data, whether it be weather data, statistics, research, geographical names, geographical information, and but even cultural heritage that's been digitized by European funds or by funds or by public funds, we are obliged to actually open it up as open data. Because I do believe that it is, I do believe that the, I really believe in the economy of commons, the commons economy, in that when you share something, you always get a lot more back than, you know, of course there might be instances where people abuse data and open data, but I think the benefits will always outweigh the risks that you're taking. And there's a couple of more facts that I want to say about linked and open data is linking data is not possible without open data. That's something that we really, really should be very clear on because a lot of people say, oh yeah, we're doing linked open data, but there are data that's not open, so you're not doing anything, you're not contributing to the whole ID of linked open data. Secondly, of course, open data doesn't necessarily mean that it's open content. For example, you can publish the metadata of a documentary, it doesn't necessarily mean you can actually show the documentary or push the content, but copyright shouldn't be a showstopper, you know, look at what you have and what you can actually publish and you'll see there's even bits and pieces of information of artificial assets that you may not be able to publish it, but you can actually publish, you can publish the metadata, so open data, and that's already a huge, huge, that's already important. Why do I think that GLAMs, the Galleries, Libraries, Archives and Museums should lead the way in this whole linked open data movement? It's because, first of all, they already have an incredibly rich and structured datasets, which they've accumulated over the years. We're really good at that, that's what we've been doing. We actually have experience to reach out to audiences, and basically that, for example, also it enables us to carry out evaluations with end users. We're excellent metadata managers, we have a lot of knowledge on a wide range of subjects, but as I said, we're not the authority always, but we do have authority and knowledge. Let's face it, we have always been trusted repositories for government agencies, professionals, artists, I mean, whoever, even in a commercial setting, whoever creates information, knowledge and facts has trusted us to actually keep their stuff safe, and then there is the principle of lots of copies keep stuff safe, so that's also the reason why I think that the GLAMs should really lead the way. It's also in their own interest, as they have a lot of limited resources. I won't go over this, but this is just a list of linked open data initiatives and projects that the Netherlands Institute of Sound and Vision has actually worked on, and one of them that you see is the network war resources or the network war collections, which is more recent. I created a tiny URL which has a lot more information on those projects with links to all the demonstrators, the tools, the applications, and the papers that were published as a result of it. What these all have in common is that they're very academic, and where I do think what the flow was very often is that there wasn't a clear use case. They weren't real end users. It's like, you know, obsessed by your customers, that's really what you should be, and I think that's why the network war collections as a project has actually a lot more longevity than its academic predecessors, because what they do have in common, these projects, is that they collaborate, whether they want it or not, you know, whether they want to or not, but collaboration is really a key issue. Collaboration openers are really the main drivers of the success of the open data initiative. And the use case that was missing in the more academic linked open data projects is really something that's super important in network war collections, because we brought together an amazing group of museums, collections, archives that have, as a main core, as a main mission, to keep the memory of the world war alive. As you all know, the last eyewitnesses are dying. So basically, how are we going to keep the memory alive, just to make sure that history won't repeat itself, that people just won't fiddle around with history and just make up a whole different story. And I think what we've done with network war collections is really get that group of people together who are actually convinced that this is what we should be doing, and that what we really achieve is that we want to use all these sources, all this information that we have, bring it together and tell maybe not a new story, but a more reliable story, and gain new insights in what happened to victims, to aggressors, to events in the Second World War. So where are we at now? We start in 2016, and so far we already have more than 86 different museums, institutions, and so on, who have published their collections as open data to this project, which is quite a thing, because to a lot of them it was very, it was quite a threat, because opening up your collection means you're opening up your biggest asset, what was going to happen, and actually they've all come back to us, saying that they were super enthusiastic, because now they know more about their collections than they ever knew, their metadata has been enriched. And so basically how we do that is something that the Data Evangelist, that's how I call her, Lizzy Jungma of the project, where she will explain how we all go about and where we are sending out. Thank you, Tom. I would like to start with quoting Mark Twain, who said that history doesn't repeat itself, but it does rhyme, and that is what is important for us. We don't expect a new world war, but there are definitely lessons from the past that we can learn nowadays. But how do we try and educate people about the second world war? Well, this is a website, and you can say this is just an interface, and it is, it's a Dutch interface, so I try to use Google Translate to translate it, but it can still look a bit disappointing. Whereas the big linked open data system that I expected you to show me today. Well, even though we currently have 11 million objects in our collection, and it's 226 collections, and we even have 1,300 films, we tried not to be the linked open data thing that most people expect. This is what I call the Sparkle Endpoint tyranny. We work a lot with people from universities, researchers, and they are all linked open data evangelists, and then they collect all this data, and then you ask them, well show me your data, and this is what you get a Sparkle Endpoint. And I was so happy to hear a lot of talks this morning about your users. I'm a trained historian, and I know my colleagues, and none of them is able to get anything out of this. So, what's linked open data if you can put it into a system, but no one is able to get it out of that system, because the interface is super complex. Well, it even takes me a lot of time to get data out of the usual linked open data system. So, let me go back to this picture. Everything is linked open data except for the interface. I'm totally trying to avoid difficult interfaces. I'm trying to use the ideas of linked open data in a completely different way. So, what makes us different as a network for war collections? What makes us different from other Glam Institute? First thing, we don't own physical collection. We don't. People keep on offering us physical collection, because they've seen our portal site, and they keep on calling me, you know, I have stuff from my grandparents. Do you want it? No, bring it to a museum, bring it to an archive. Once it digitizes, I want to have it, not the physical objects. So, we're only working with digital information, no digital, no real objects. But what we do is we help in red the content department. We help Glam digitize the collection. A lot of the organizations we work with, they have four people on staff. One is both the IT person, the morning manager, and the volunteer guy, and is also the person who needs to do all the digitization. So, a lot of organizations we work with, very small organizations, little knowledge about IT. We try and help them digitize their collections. And we also try to help them publish their digital collections online. Plus, we advise them about all the rights and legal aspects of their data and their digital objects. Because one of the big mistakes, and I guess you all noticed, one of the big mistakes people make is that they think that if the object that they digitize is copyrighted, that the metadata is copyrighted as well. So, we get weird debates about copyright claims or metadata, which is not true. As an institute, when you create metadata, your metadata is either within your copyright or it's copyright free in the Netherlands. So, you know, skip the debate. That's one thing we do. So, we help create our own collections. So, we have a very personal interest in it. We need the digital collection. So, we're interested in organizations digitizing it. But we're also creating reference data. The who, the what, the where, and the when. And this is my next slide. What we try and do is not just ingest everything about the Second World War. We also try and explain something about the Second World War. If you put 11 million objects in a database and you put a Google box in your interface, then you're not helping people. You really aren't. It's like what Google itself, you Google for something and you get 100,000 hits. And Google is good in estimating what is relevant for you or a goodish. Well, as a historian, I know that it doesn't always understand what I'm looking for. But we're not as good as Google. We don't have the kind of big data that Google has to make these kinds of assumptions about a user. So, we don't have reference data to, you know, make all kinds of automated guesses. We need reference data to create our own estimations. So, this is what we do. These are the four historical questions every historian asks. And we created, you used the word lexicons this morning. We've called them we created for these four X's. And it's not this side. I'll get back to this. All right. I will show them what they look like. But this is what we create. And the network only has four people on staff, four and a half. So, this is all the time that we can invest in creating something that is understandable for huge audiences. So, what we do is this is in our kitchen. This is what we do in the back end. So, most organizations send us data, their metadata, in one form or the other. Thank God a lot of vendors now have APIs. So, we can harvest APIs. But a lot of times people just send us Excel sheets. I even got USB sticks by mail with handwritten notes to it. This is my collection. Can you use it? So, one of the basic principles of linked open data, of course, is that a collection owning organization publishes its data as linked open data on the internet. And I'm able to connect to it and, you know, voila, it'll work. But the reality is that most museums don't have triples, don't have API, don't have anything except for Excel. So, a lot of times we convert metadata in any form to XTLT and we create triples for organizations. This is not something museums are currently doing. We're doing it for them. And it's not very difficult when you know what you're doing. We use very simple ontologies. We use Dublin Core, we use schema.org. I've seen tons of theoretical ontologies created by every research group in any university in the world. So, I always check, like, who's maintaining this ontology and has it been updated since 2002? So, and then you end up with Dublin Core and schema.org. That's the reality. So, you could, I'm doing this on a notepad and it works brilliantly. The output of the conversion first goes into our system. This is a screenshot and we use a system that is very visual, thank God. But as you can see, it's not very, it's not very linked. It's just a screenshot of a lot of collections, different collections just sitting there being a collection. There's no interaction yet. Although, at this point, I can already send triples back to the museum or the organization that gave me their data. So, we can send them RDF files or N turtle files or whatever they want. So, they can then use it within their own systems as triples. Unfortunately, most museums are clueless what to do with the files I send them. So, I still need to find a better way of sharing our data with them so they can actually use it and not, you know, start out crying in tears when they see what I send them. But again, this is just, you know, different collections in the system standing there. What we then do is we use our scissorize. So, these are the, this is a screenshot of one of the scissorize I talked about. We have a couple of them. This is the, what is this scissorize? It's about object names, object types, but it's also about historical events during the Second World War. And it's also about places like concentration camps. Currently, we have over 5,000 keywords in our scissorize. So, ranging from object names to, I think, can you see, 431 European concentration camps. We also have 160, 170 camps in Asia in the system. But we also have a scissorize for notable people during the Second World War, consisting of over 1100 names. And we harvested wiki data, the geo section of it, to detect places in the objects that we get. So, how do we use this? Well, first thing is, we can plot camps, for instance, on maps. This was the first thing that we built, simply based on our camp scissorize. We plotted a map with all the concentration camps in the Netherlands. The cool thing was that immediately people from wiki data started contacting us and asking us, can you give us your data about camps? We're lacking half of them. We don't even know where you've got them from. And this is, one of the things we did together with wiki data, and this is one of the collaborations I'm really proud about, is we very easily were able to share all the data about the concentration camps we have with them. They matched it with the concentration camps that were already in wiki data, and they gave us their queue code, the queue codes are the wiki data identifiers, they gave them back to us so we can now reuse them. And we know when we talk about the same camps or when we found the camps that they don't know about. And on the other hand, when you publish your camps on wiki data, you can see that people are using them. Some of the camps did not have a description on wikipedia yet, so most of the camps now have a description because someone picks it up and starts editing and starts adding information. So the outreach on our website is really nice, but the outreach on wiki data and wikipedia is mind blowing. So that's one of the nice things about open data is that if you have it and you share it, someone else can do really cool stuff with it. But as I was saying, our collections were just sitting in our system all by themselves, all by myself, doing nothing. So you want to do something extra with them to make them link, to make something completely different. Because a collection, you know, if you just publish it in your portal site, it's the same as the collection on someone's website, but then with a lot more objects standing next to it doing nothing. So this is a software that we use. It's built in the Netherlands and it's, for me, this is when, you know, elastic search nice, but this is really cool stuff. What it does, it splits searching up in building blocks. It's like playing with Lego as a search engine. So what I can do is I can get a collection and I can extract all the different parts from a collection. I can extract classes, I can extract text, I can extract data from it, and I can start matching stuff. So in this case, I match one collection, the image bank of the city of Nijmegen, and I can compare it to the WO2 thesaurus, but I can also match it to biographies from the Second World War. And what I can do is I can look for, in the collection, I can look for descriptions, I can look for subjects, I can look for titles, I can look for dates, and then I can start comparing it with other collections. And every comparison leads to a new set and leads to connections. For instance, I have a keyword in Nijmegen which matches to a keyword in my biographies database, and I can say, okay, this is the same as that. And the nice thing is that you can also use it to compare less obvious things. One example I always use is that in the Netherlands, we started using word bunker after the Second World War to honor the Americans for liberating that. Before the Second World War, we had a different word for bunkers, a typical Dutch word. We can use it as alternative terms, and we can make comparisons and still, if someone uses the Dutch word or if someone uses bunkers, if someone uses plural or singular, we can all match it and it all ends up with the same keyword. So this is one way to connect all these collections. And then we give up the notion of collections. So our second step in processing data is we still mention with each object what collection it comes from, but it's not very relevant in our context. It's relevant for the organization that contributes the data. But when you publish a report about the Second World War, it's only of secondary relevance where the data comes from. Our audiences are interested in photographs, they're interested in films, they're interested in the battle of R&M, they're interested in D-Day. They're not interested in the city hall archives of the city of Nijmegen. And I remember that this was a big shock in the beginning. Every organization I worked for was always, you need to put everything about the organization on your homepage. People want to know what department you have. They don't give a fly. So really, I worked in the arts and crafts of the 19th century department. Most people can't even spell that properly, let alone search for it on your website. So give up. So this is what we do, we try and figure out what people are interested in and then try to cluster all the objects that we have around their interests. And then, this is what we do with it. And we're going to redesign the website using users because this was only the first iteration. Now people are actually using it so we can still, we can see where the flaws are. But one of the cool things that we do is, you can see the red tags here. This is all machine done. So this is the output of the process that we went through. This specific object from an image bank in Samstatt describes the liberation of this town. But they did describe a lot about this town, but they forgot to add keywords. So, you know, when you look at it, in the beginning, there's nothing to hang on to, to use this collection. But when we started using our systems, we could add keywords to their objects. And we can also make the distinction between places and people and other keywords and we can share it back to them. But also, it makes it a lot easier to find this object. And if you find this object, you can click on these keywords and find the relevant objects from other archives, like books that are not from the same city or monuments about the same event. So we created a Thesaurus and it's also our own encyclopedia. So, for instance, if you search for the concentration camp, mouse, house, and really awful concentration camp, we don't just present documentation about it. We also write very small articles, bit like the Wikipedia header, about these camps where we get it from Wikipedia. And we also add keywords, relevant keywords for this camp, so you can, if you don't know much about the concentration camp, if you go, if you get this facet, you can still click and find all kinds of different information about this specific concentration camp. You even get people that were incarcerated in this camp. And you can find a mark on the map where this camp was located. So this, again, we're going to redesign the website this year to even make it a more of a user experience, help them search and research the Second World War, even if you don't know much about it. But we're all, we are also doing a different project called War Lives and that was because one of the things we found out in our statistical data from our website was that most people that are searching our collection, they're not searching for the battle of R&M or D-Day. They're searching for the grandfather, their ancestors, their nephews, maybe people they knew from the Second World War. And we didn't have much information about people in our collections. Actually most archives have information about people like death registries, marriage registries, birth registries. But it's, they have it in separate systems, usually for people doing ancestry research, but they don't really use it for historical presentations. So that's one of the reasons why we didn't have much information about people and we started actively collecting information about people and we used linked open data to reconstruct lives. For me it's very logical but I know a lot of people are surprised what we're doing with people's lives. So this is a screenshot. We're going to publicly launch this website by the end of this year and we could launch it now but unfortunately Europe has very strict privacy laws so we have to prove that everyone that we publish about is dead. Well I can tell you they are but you know proving it in a legal way is still a difficult process so we're in the middle of the process getting it cleared so we can publish it. So what's next level about reconstructing people's lives? Well first of all because most archives aren't very structured in annotating people's lives or annotating people. We had to find many different resources about people that participated in the Second World War. So it's not just archival documentation. We also had to crawl a lot of websites, mainly commemorative websites. In the Netherlands there's the Jewish monument about all the Jewish victims of the Second World War but there's also you have also have local websites like this one. It is about the people that were picked up by the Germans during Arassia and were all deported to Germany. A lot of archives have archival material that isn't digitized yet so we either crawl websites or we help organizations digitize this kind of material. We do anything to get this data and again we confer people's names to link to open data. Again same tools, XSLT, notepad. What was interesting was that we couldn't use Dublin Core to describe people so we had to make up our own metadata schema or actually I didn't want to create my own ontology. I wanted to use existing ontologies to create information about people. So what we use is a lot of schema.org and only occasionally do we create our own namespaces if there's nothing in schema.org that I can use. What I learned was that you can best split a person's name, create a sort of metadata about a person, name, place of birth, place of death, date of death, date of birth. That's the metadata about a person and in all these resources you can find information about people in a specific event. For instance the transportation archives of the Jews being deported to Germany. There's a lot of archival material about it so if you create a transportation event then you can try and match names of a person in the transportation archives to a name of a person in another archive for instance a birth or death archive. It took us a while to get this right but it's really cool and this is unfortunately very light. I'm very sorry but this is where the building blocks that we use are really cool because this is where we can use all the search algorithms that you know you have at your service. So we use everything both the Eierowinkler and Levenstein and you weigh and you match because in most archives people's names are misspelled by one or two letters so you can weigh that. Dates of birth usually have a date of birth but sometimes they were misspelled as well or people lied about the date of birth for instance because they didn't want to be deported so you need to take all these variables into account and do a lot of calculations and then you can match names of people from more generic resources to very specific archives about specific events and this is what we now have so we have entered 1.2 million events and birth and death are events as well and we were able to reconstruct the lives of 230,000 people so for me there's a big difference between a name and a person a name is something for instance the writer of a book you have a name but you know nothing about this person until you start accumulating data about this person and that's what we're doing here out of all these events we try and recompose a person's life and this is what a person's life can look like in our own portal site so we try and explain as much as possible on a timeline about this person so it's a bit like the horrible version of Facebook where you have a timeline of a person's life this person was transported a lot of times to many different awful concentration camps and because we collect concentration camps we were even able to plot his personal route on a map as you can see here and the thing is we were perfectly happy about it and we showed it to people that you know we're googling for names of family members and they are in shock because a lot of people don't really know what happened to their relatives they know that they're dead but most people don't know how much happened to their ancestors during the war and that is one thing that we try and visualize is what could happen in a war and of course you know you start doing big data stuff we we don't publicly do this because you know this will give us a lot of this debate but we plotted the deadliest dates and the deadliest places for Dutch people during the Second World War on a map and a timeline and I won't go into it that much but this was very easy to do and it also shows that there's a big misconception in the Netherlands and that was for instance that the beginning of the war 1940 was a very deadly period which is not true and another assumption most people have is that during liberation and the hunger winter in 1945 a lot of people died whereas actually most people in the Netherlands during the Second World War died in 1943 so this is all the the powerful things you can do with these kinds of with this data and we're still exploring it but where do you draw the line and I showed this picture to Tom and he was looking at me and saying okay that's a nice picture what is your story here well as you can see the first person was a young boy who died before he reached the age of 10 the second guy is a very sad story his brother was the national socialist movement leader of the Netherlands and he was executed because his brother was the leader of the national socialist movement third person also died very young but the last but he died in Russia so he joined the SS and started fighting at the Eastern frontier and we had we still have a lot of debates do we mention these people in our war lives website or is it a commemoration site where we honor the victims last person is even more interesting Katja Schot was a camp commander and she was awful she killed 10 women because she had a bad evening and she was convicted after the Second World War but she's she's still alive her image is published on many different websites because there are images but usually they don't explain who she is so you just find her image do we take her into our system and where do we draw the line I think we should include everyone in our portal site but I'm very much aware that with the last person we will also create some sort of new witch hunt for her I know she's already mentioned on the website looking for Nazis in the Netherlands and so as Tom started his talk we are trying to fight tyranny by showing it so we're trying to show what it looked like during and before the Second World War in Europe and we hope people can use all the data that we have about the Second World War to study the Second World War and to interpret things that are happening in our days and we publish it on our own website and we hope that people we are trusted because we collected trusted resources but we also share everything that we collect with everyone else one of the projects we did with sound and vision was that we researched their collection and we found Nazi propaganda movies that were out of copyright and we published them in collaboration with sound and vision again on wiki media comments so everyone can watch them can use them to write articles on wikipedia or and can use them for whatever but just showing these images we think is helpful to understand both Second World War and our current times so I would like to end with this last movie and this one is really important and for me really important to share because this is the only filmed footage about the show this is an actual concentration camp this was filmed by a filmmaker in concentration camp west of war and this is the only proof that people were really deported to concentration camps so could you please show this movie and as you can see Germans were administrating everything and we're a bit thankful because we can use these administrations to reconstruct lives a couple minutes over but I'd like to entertain a couple of questions about that extremely powerful presentation thank you Lizzie and Tom I have an immediate question I think about things about the fact that you know the current American administration with some of the agencies in its service kidnapped thousands of children from their parents and we now know that that number was actually much higher than was reported a year or two ago there are thousands of children who are still disconnected from their parents and I think about utilizing both the kind of schemas and technologies that you guys put into work here for an historic event have you done any work with activist organizations to encourage them to utilize these types of technologies much closer to the occurrence of events so there's a richer history both immediately accessible to us as we do research as to what actually occurred and how it affected even individuals because it seems like a very very powerful set of approaches and technologies to to bring to solving more immediate issues no we haven't I hadn't even thought of it I think that would be a very good application we do work with a lot of historians because the problem that we ran into when you have a reference to a name in an archive and you wonder is this the same person that is referenced in another archive most historians deal with this problem so we work a lot with historians and they all recognize this issue and next month I'm going to do another workshop with a lot of historians but we're looking into the golden age or you know 18th century archive for documentation but I think current in a current situation could be another very good application yeah I'll mention we had Yvonne Eng from the witness foundation speak last year and they do a lot of work on tracking media for contemporary events and human rights abuses and I'd say witness would be a fantastic organization for you guys to consider working with because you know they're they're studying a lot of these types of contemporary events that are captured in rich media and I think these approaches would be very powerful when helping them in their mission