Jason Evans 0:03 yes so yeah thanks for inviting me to share a little bit about our, our work with IIIF this afternoon. The National Library has been working with IIIF for some time and but my talk today is going to focus on how we've been using it to help with our crowdsourcing activities, and in particular how we're using it in combination with open data and Wikimedia projects in particular so we're talking Wikidata and Wikimedia Commons. So, this is the IIIF viewer the National Library of Wales uses. We've been using this since about 2016 for pretty much all of our digital content. And this allows us to annotate images to create galleries and multipage items and to display data, all the wonderful things that IIIF does so well. At the same time as this investment in IIIF the National Library has been adopting an open access policy and sharing digital images and data from our collections openly on other platforms, particularly Wikimedia, and we've shared around 20,000 images to Wikimedia Commons. By this point, and a lot of those images also have Wikidata items about the items that uses our metadata from our collections. And although Wikimedia Commons has no official support for IIIF yet, and there are discussions and talks about how IIIF can be used on the community comments. And there are some really interesting tools being developed around IF in the Wikimedia space, so I'm going to talk a little bit about those. So Wikidata. For those of you that aren't familiar with Wikidata. It is essentially a linked open database that anyone can edit contribute to and reuse, and it contains data on just about everything. And a couple of years ago, IIIF there was a IIIF manifest property created and that allowed people to share a triple IIIF manifest relevant to the item that's being described on Wikidata. So this is a really interesting development for us and we immediately went about sharing manifests for lots of our images that we have on Wikimedia commons and everything, as I'm sure I don't really need to say about Wikimedia projects is massive. So there are millions of articles on Wikipedia there's millions of images on Wikimedia Commons and and hundreds of millions of pieces of data in Wikidata. So naturally, this is a very stable platform, and there's a big community developing tools around it. So it's a really interesting space for IIIF. So as an example, one of the kind of backbones of Wikidata is its query service. So this allows you to query, all of the data that's held in Wikidata, and of course that includes large amounts of data for collections in archives, libraries and galleries all over the world. So in terms of the context of a national collection. I think it gives us a really clear example of how it can work and how we can draw all these collections together. So using the query service, we can ask for all sorts of subsets of data that include IIIF manifests, so they could be collections from one institution from multiple institutions with particular themes or related to a particular location so we could ask for manifests for collections that are located in Wales, or the depict the south of France, there's, there's kind of very little limits on how you query the data. So, as a good example of what kind of interaction. This allows within literally within days of us sharing manifests, to Wikidata for a lot of our images, an Italian website that aggregates cultural content, used Wikidata, and that query service to pull out all our content that had IIIF have manifests. It then used those IIIF manifests to add 1000s of our images to their website and you can see them here. They've pulled in, still images, and then you can click on those and see a zoomable version of the images, and they pulled in relevant data like titles and authors and dates etc, from those manifests. That's a really good example of how IIIF and Wikidata can be used to share content across multiple platforms. Recently, we worked with a website called watercolour world, and they approached us and said, Can we have watercolours from your collections on our website, and literally what we did is we queried Wikidata for watercolour paintings in our collection that are on Wikimedia Commons. It gave us a list of manifests we gave them the manifests, and they did the rest they were able to get copies of the images, and all the relevant data from that to upload their website so it made that collaboration really easy and kind of low on resources for for both parties. And then there's the Wikidata community itself which has been starting to develop tools around IIIF. So, this tool was launched a couple of years ago, this is a tool for tagging, the location in an image of the pic statements in Wikidata so if you have a painting and Wikidata, it might have a statement saying it depicts trees or house, a particular person. This tool allows you to then tag where in the image those things here. And it stores that data as IIIF coordinates. Back to Wikidata automatically. The tool also allows you to add additional statements to pick statements to Wikidata, and it allows you to generate manifests that include the data that's just being captured. So we've been using this as a crowdsourcing. We had a bunch of volunteers that could no longer access our building to do research to do volunteering tasks. So we gave them a big long list of images that we have on Wikimedia Commons that have pic statements on Wikidata and said will show us where they are tagged tag the images using this tool. And we've tagged 1000s of images during the project. In this way, and that links nicely to another tool that's been kindly, developed by a wiki volunteer. And this tool allows you to start to explore the fragments that have been captured by the previous tool. So, where people have tagged things in items you can search all of those tags elements by using the pic statements. So, we can search across collections or across individual archives. So for example, I was able to use this tool to find all the coats of arms that have been tagged in our images, and wonderful selection of hats. So, this is a really, for me, although these tools are basic. They show what's possible and what kind of tools we can develop and to enrich our user experience on our own website in the future. So the next step for us was to incorporate our internal use of IIIF with the open data that we seem to enrich our own collection data. And we have recently developed a new crowdsourcing platform, which uses IIIF to pull images in and allows volunteers to annotate and describe that content using IIIF. So what we've done for our project with a series of photo albums is giving people the ability to tag where things are in images, rather than simply tagging as free text what appears in the images we've used Wikidata so it connects to the Wikidata labels, so people will tag. The item for a bridge in Wikidata, and we store both the label and the Q ID in our system. So, that has a number of advantages over just having a free text, sort of way of people describing images. So, firstly, it gives us a standard vocabulary so every time someone tags a bridge it's going to be spelt exactly the same way it's going to appear the same way with the same identifier, and doing it this way allows us to access all the other data that's in Wikidata, so we can access coordinate data we can access images, links to Wikipedia, we can access identifiers for other institutions that hold relevant information. So we can access all that extra crowd data that's there on Wikidata already because Wikidata is multilingual, we can display in multiple languages the data that's being collected, we can display it in French and Chinese in English and Welsh. It gives us a lot more flexibility in that respect. And as a bilingual institution, it means we can also allow our volunteers to tag in multiple languages. So, if someone wants to tag an English they can search for bridge. If someone wants to tag in Welsh they can search for bond, the Welsh word for bridge, and they will be identifying the same item and tagging the same entity within that image. Of course this also means that if we want to share this content to Wikimedia at a later date, we've got all the Q IDs for all the things that are depicted, and it makes it a much more streamlined process to push this content out on Wikimedia, where it's highly visible so this conundrum that I know I'm sure we're not alone as an institution in facing is what you, what do you do with all your crowdsourced data. This gives us a very easy way of putting all that crowdsource data on Wikidata where it's highly visible and easily accessible by anyone who wants to use it. So, basically, in my opinion, IIIF, and Wikimedia, have a lot of potential as as partners, they have the same kind of goals in terms of improving access to information, and there's so much data out there already, that can help enrich work that's being done through IIIF, and I've shown you some great tools that are already being developed in this space, but I think there's a potential for a lot more and I'm hoping as more and more institutions adopt IIIF, as a standard, that we will see more and more tools being developed in this space. Thank you very much. Transcribed by https://otter.ai