 Y gynllun â'r gweithio a rwy'n meddwl i'ch meddwl am y gwasanaeth y gweithio ac yn ddionio'r yma. Rwy'n ysgol i'r OER ac eu mwyodraethau, yn meddwl eu cynllun o arferwyr yma, eraill o'r blefau Chymru, a yw'r unifredigfeydd i'r University of Oxford ac yn unedig gwagwr gwahanol oherwydd rydw chi sut ichi wneud bobl i'r meddwl yw methau Chymru. Yn ymddangos, Martyn Porta! Os i wnaeth yma i Martyn. Mae gweithio'n baith. I will start my talk by going back historically to how I see the OER, what we did in the OER movements for having been involved in the funded UK OER projects. So there is repurposing existing material, getting material that existed and finding the people who owned it and asking could we re-license it please, could we put it on a different platform, could we have permission to remix it and make open materials with it. There was funding creation of original material that was born open and trying to kick academics off in the practice of making open materials, the efforts to shape top-down institutional policy and make it more friendly to creation of open resources. There was the dissolution of the barrier between the platforms used by academics and the social media platforms used by the Horry Ploy that people did more courses on blogs and flicker and used those kinds of platforms. And then there was the shift from the emphasis on resources to practices. So if we have the values and practices of remixing of resources embodied in communities and in people involved in learning, then the resources will come out of that as a natural process. I want to talk about, I want to say opening up another front but maybe the military metaphor isn't apt, we're trying to do something to benefit to everybody but the last session talked about an army of peace so maybe it is but I'll say opportunity. There's a huge opportunity I think additionally to other things in the re-purposing of data to make educational objects, data that was not intended to have an educational purpose initially. So examples, the data describing the collections in a museum or a culture institution. There's data from the outputs of research projects, so this is an archaeological project that surveyed archaeological sites in the British Isles. There's data from the public sector, not that the fertility rate of a particular country in a particular month is a learning object but more of that data, different types of it, given context put in some sort of interactive platform, make something like GapMinder, something that people can come and interact with and can support learning. So the same is, I'm not expecting the computers to do the teaching, I'm talking in this session about resources and ways to ease the production of resources that are interactive and interesting and data bring with them their own problems so that something is called data and it's in the computer and it's official, it can have a seeming authority to it, which of course data aren't always authoritative or aren't always neutral, we have more data about certain parts of the world than others. So creating objects from data doesn't solve social or cultural problems, it brings the traditional problems associated with claims to knowledge and other problems as well. But it's still, we can make things that support a process of learning and make it more interesting and we can make those things more easily than before. But why in particular now, we've had databases for a long time, we've had websites driven by data for a long time, we've had data that's freely accessible and data that you can get through an API for a long time in all these categories. Well there's a problem, a deep problem of communicating between databases and joining them up. On the screen are the different identifiers in some database for Jane Austen. So she's, for New York Times she's person slash Jane Austen for the women writers database she's 61CA blah blah blah. So they don't represent the same object in the same way and if we look at the say the relationship of authorship, the relationship between a work and a person of authorship, that'll be represented in a completely different way in a library database, in a New York Times database and so on. So the existence of data about different types of data about the same thing doesn't mean we can join up and create something that explores knowledge about that thing. We were told that, yeah I'll move on to the next slide, so if you're going to make an application like this since you're in a cinema, I've chosen one based on films, this is a fun application for exploring the relation between books and films. So you can go into say Les Miserables and Les Miserables is a film based on a book, a book by Victor Hugo, Victor Hugo also wrote this book and there are lots of films based on this book of Notre Dame and you can read the text of the books through the Bibliotech Nacional Interface. So to do this traditionally would be a lot of work because you've got to model your domain. So how do you represent a person, how do you represent a film, how do you represent a book, how do you describe the relations between the director of a film and a film and then you've got to enter the data and then you've got to think about how you'll present it, you've got to make this site which is an interface to what you've made so that would be the old way to do it. And we were told this problem would be solved. In the initial design for the World Wide Web there was this idea of the semantic web that there'd be standards and formats through which databases could talk to each other and say I have information about this thing and that's 20 years ago, no 30 years ago it's easy to think that that is just hype or vapourware because it was very hyped that it was going to happen and it didn't happen. But I think it is starting to happen that standards are being adopted, new databases are being created with, built with a kind of an open format and kind of built to share data with other databases but a crucial step is the creation of a hub. So back in 1992 Tim Berners-Lee said about the World Wide Web that the number of documents people could get through the World Wide Web would eventually be absolutely astronomical. People would need to manage that explosion of information and one recommendation he made was what he called the Exciclopedia. So there'd be this site continually updated that you could go to to get for any topic in arts, sciences, history, culture, an overview of that topic, an introduction and some pointers to the key research, the key sources in that topic. And there were several attempts to build the Exciclopedia that Berners-Lee described and one of them happened to be hugely successful and became Wikipedia which is kind of the first point of call for all sorts of categories of user. So the crucial step for structured data for the knowledge that's not in a narrative but in the form of a structured database is having a hub and that the newest sister project of Wikipedia is Wikidata which aims to serve that role, a summary of knowledge and culture in structured data but also a gateway to other sources of knowledge about that same thing. Yeah, some stats about it, so it's five and a bit years old with currently 46 million things described including more than four million people. These numbers are already a bit out of date because it's very rapidly growing but it's already the biggest digital authority file for people ever created. It's already very dense with say geographical items to have a co-ordinate location but it's an inequality that's not equally all over the world. So this can be a kind of central reference point so I'm going to go through the slides I've shown you backwards. So Wikidata exists, it enables, it has its own visualisation tools and its own ways to get data in and out and it enables, it solves some of this problem, it enables to create views of a particular domain of knowledge. So this application is actually built in a hackathon using Wikidata. Okay, so the IMDB have a database of film posters, Bibliotech Nassanal have an online access to texts, Wikidata knows which is which, knows what the links are to the internet movie database and to the Bibliotech Nassanal and has those based on relations or has a lot of them if you spot some that don't exist you can add them. So what the developer of this has to do is just present it, just how to work out how to present this to use and how the user interacts with it. It cuts out a load of the work. Wikidata has thousands of other databases reconciled to it. So one query gets all of these and these aren't just identified, these are links. So to say the information about Jane Austen's grave is at the find a grave link number 44. This I put up, this is UNESCO sharing data. Wikidata isn't ideal for time series but they can be put, so there's lots of like time series like the population of London month by month or economic data month by month. Not all of that goes into Wikidata but it can go to somewhere that's addressed by Wikidata and we can make gap-minder like things. This is a research project in Oxford that I've persuaded to share not the output of their research but the surface layer of their data. What they find out about where was it, what's it called and what type of thing is it and then we can generate gaps, different visualisations of the data and as well as urging developers to do this I've had to kind of eat my own dog food in sharing museum data which I'll show you briefly. So we shared a load of data from a university collection. Those data are ostensibly about a bunch of items and those items are made of paper or they're made of wood or bronze and they're 20 centimetres high or 40 centimetres high from a particular place. But implicitly the collection data of museum is about other stuff. It's about Krishna, Buddha, Ganesha, it's about Mount Fuji, things are depicted, things that people made these artworks for for the purposes of or to depict. So we can give access to the collection in terms of those things. We can use the things that the collection is about for discovery of the collection and we can get context. So this is the plan of the different eras and dynasties and cultures represented in the collection and because it's all one database, bibliographic, biographical, geographical, we can draw in contextual things, we can draw in other names for the item, we can draw in the extract of the Wikipedia article so we get a few sentences saying what this is and why it's important, we can get a contextual image and then build a web. So this era is connected to people, there are works connected to it, it's connected to eras and time. So people go on the journey through a culture to a person, to a location on the earth surface, to another person connected location to the works of that location to what was depicted in the works made in that location. So we can create these kind of web structured things that people can explore. So when I was doing funded open educational resource projects, I was sharing a lot of Microsoft documents, words and power points on the web and those have considered educational objects. If I'd done something interactive like that, I would have got tens of thousands of pounds to do it. So I think we can think more expansively about applications that give an overview of a domain. I've shown you films and books and I've shown you Eastern art and it's basically the same code and the same principles but we could do that for any domain and let people find their own way through it. There's a few other things I won't mention. But yes, histropedia, if you haven't tried histropedia, have a go at that which it's another Wikipedia driven application and it makes amazing interactive stuff. But finally, I just want to say that what about the path, these things are very web-like, they have multiple entrance points and astronomical number of pathways you can take through them. Who decides what path the learner should take? My earliest memories of being enwraptured in learning are staying up to watch James Burke on Connections and Carl Sagan on Cosmos and they had these narratives, the programmes that take you through the history of ideas or take you through a topic and it's one person's view of the connection between different concepts and I thought the open education revolution would be about that but taking the paternalism out. So you'd have different narratives about a topic and you'd find your own version by assembling different ways. So we could do that. Too much data is still locked away in proprietary platforms or platforms that can't be repurposed so open data advocacy should be a big part of what we're doing for open education. Thank you. Thank you very much. We're going to take questions now very briefly so if anybody has questions, a wonderful thing about Martin is you learn so much when you listen to him even for a few minutes. On the Open Education Seag webinar collection, Martin did a wonderful webinar for us which takes you through all things wiki so if that's not something that you know a great deal about they've produced some amazing things and they're very succinctly portrayed within that webinar very cleverly and clearly communicated. Do we have any questions? It's wonderful when somebody takes something so complex. I just want to ask about the last comment you made about James Burke and obviously it's been very interesting recently because kind of Clark's civilisation thing has been remade as civilisations and it's been an opportunity to really think about this whole issue of paternalism, interpretation, curation and whether the interpretation of an individual who knows their way around objects, history, the library, whether you can separate that individual from the content and whether the content can actually stand on its own and you can still sort of find your way through. I'm not sure about this. This is why I said early on that I don't expect the computer to do the teaching, that this needs some context whether it comes from formal or informal learning or whether it's individual curiosity or whether it's something you do in a community or with mentorship and whether the mentor is present or there's somebody on the TV. There's got to be some context but I hope we have an opportunity to get away from the broadcast model because we would tune in to Carl Sagan and James Burke in our millions and watch the one guy giving which was great and gave us an illuminating journey that we wouldn't have thought of ourselves but the same facts can sustain multiple narratives and find multiple interesting pathways so it's creating the opportunity for that but in a context of reliable information so I've emphasised authoritative sources of information all the way through so people as the saying goes people have a right to their own opinion but not to their own facts so taking so a kind of paternalism in kind of quality but being ditching the paternalism of the pathway what's the significance. But there is also something in the multiplicity of description of a Lord Clark or a James Burke that actually they reveal something of themselves and their own understanding of stuff. I'd like to see other people revealing something of themselves people from a different culture or a different yeah my my education was a quality education but it was from particular points of view and then adult life I'm learning about different points of view so I think that's the point you're making as well yeah. Yeah we can open up the as much of the yeah the cultural material and let people create their own stories their own narratives. Hi Martin what's the equivalent you know on Wikipedia there's the whole kind of deletest version include a version you know what's the equivalent of that in wikidata. So there is a notability criterion for wikidata as there is wikipedia it's much more lax for wikidata so wikipedia it's got to be substantially covered in three at least three reliable sources wikidata is for things that are in a reliable source I described in a reliable source and it has to be some rich description so I don't think I could put some museum catalogue record today like textile fragment from somewhere in India purpose unknown and you couldn't put that but something which has properties and exists in some official register if it's a person an exhibit a book whatever that can be in so that's why there's 46 million things in wikidata at the moment and there could easily be 100 million it's rapidly expanding whereas much larger numbers than you get in any individual wikipedia.