 Link Digital Futures Initiative and the adjoining report. It's fitting that we're meeting here in Montreal to envision a linked, open digital space for the arts. I'd like to acknowledge that this land has long served as the meeting site and exchange place for many indigenous peoples, including the Genyengahage and much later, French and British settlers and today, a broad diversity of peoples. As we embark on a journey to establish a performing arts domain within the web of data, we express the hope to create digital territories that do not replicate colonial or proprietary structures, but are rather meant to be shared and to be open to all forms of artistic expressions and worldviews, including the past and contemporary expressions of the First Nations, Inuit and Métis peoples. I'd also like to acknowledge our project partners, Culture Creates, Rideau, Atlantic Presenters Association, Mass Culture and the BC Alliance for Arts and Culture. I'd also like to acknowledge that a linked, open data ecosystem requires many more contributors and I'm glad that potential partners are in the room with us today or watching over live stream. Frédéric Julien and Beatt Easterman, our co-authors of the research report we're proud to present today. It's titled, A Linked Digital Future for the Performing Arts, Leveraging Synergies Along the Value Chain. Beatt Easterman is deputy head for the Institute for Public Sector Transformation at Byrne University of Applied Sciences where he leads the data and infrastructure unit. He specializes in research and consulting activities related to the digital transformation of the public sector, including open data, linked data and crowdsourcing. Among his areas of interest are the heritage sector and the performing arts. Frédéric Julien is director of research and development at Capocoa and project lead at the Linked Digital Future Initiative. He is policy analyst, a communicator, an activist for the arts and culture sector and his interests include partnership development, cross-sectoral collaboration and well-being and research policy. So over to the two of you. Thank you, Aku. So why this Linked Digital Future Initiative and the ensuing report? Well, it doesn't come out of nowhere. Over the last five years, the issue of performing arts metadata has emerged as one of the most pressing issues for the arts sector in Canada. Why? Because the web has changed the way that algorithms work have changed. So today, an algorithm will no longer want to deliver you search results. They want to deliver you recommendations. And for this, they need data. Consequently, this sector needs a data strategy. They need it. We need a data model and we also need the actual data. Purpose of the Linked Digital Future Initiative that we are also launching today, along with the report. So this initiative is meant to first provide a data model for the sector, which comes along with today's report, and also prototype so that we can harvest data and feed it into structured data for algorithms. Of course, this isn't the whole thing because if we stop there, we'd just be talking to our small circles. So we'll be traveling Canada in the next months. To deliver a round of professional development sessions and presentations to explain these concepts to the broader performing arts sector. So that's it for my intro of the Linked Digital Future Initiative. Let's dive into the report itself. So yeah, welcome everybody. I would like to talk to you today about the Linked Open Data Ecosystem for the Performing Arts. I'm going to talk about its overall vision, how this vision has matured over time through various projects, about the elements we have been working on in the course of this first phase of K-PACOR's Linked Digital Future Initiative, about the current state of implementation of this ecosystem, about the use cases of the Canadian implementation partners, about some of the challenges ahead, and I will finish the presentation with recommendations made by the advisory board. But before I start, I would like to make two caveats. First of all, my presentation will be about Linked Open Data, but you won't have time really to dive into what Linked Open Data is during today's presentation. What I need you to retain, however, is that Linked Data is about interlinking data from databases that are managed by different institutions. Second, in my presentation, I'll give you a few glimpses of the report we are presenting today. However, it won't be possible to cover everything, so I would kindly ask you to also visit our website where you can find the entirety of the report and where you can also explore our data model. But let me now start with the vision. Imagine one big database that serves the needs and interests of a variety of stakeholders related to the performing arts, as you can imagine they all have different usage scenarios in mind when they think about data. Obviously, we're not suggesting to build one big monolith, but we're recommending to build a database with a distributed architecture made up of many smaller databases that are interlinked with each other. Before I get into the details of this concept, let me tell you a bit more about the journey so far. I will tell you the journey from a very personal point of view, so you know where I'm coming from. So my first encounter with theater data was through the Swiss Theater Collection who was looking for a replacement for their existing database. The person in charge was looking ahead and was interested in learning how databases would look like in the future. As a regular contributor to MusicBrainz, he was well aware of the opportunities of that link data and crowdsourcing, collaboration online would bring about. One of the peculiarities of the Swiss Theater Collection is that it has a large performance database covering many decades of Swiss theater. This database serves as a finding aid to the institution's entire collection. Now after sketching out the contours of the new platform, we started to write applications for project funding. Some of them successful, some of them less so. We also started to hand out some of the conceptual tasks to our students at Burn University of Applied Sciences. Finally, the project was officially launched and we were able to document the requirements of the new data platform in more detail. 2017 was then a very eventful year. It was the year when we started to internationalize the project, first in the form of an application of a European project. Unfortunately, the application failed but we made enormous progress on the conceptual side. It was also the year when we published the first draft version of our data model and started to implement parts of it on Wikidata. 2018 then brought a completely new turn. I got involved in the project focusing on interpretation research on the basis of a large collection of audio recordings from opera in New York. Of course, there was a lot of metadata going with it but more importantly, there was also the new perspective of researchers and the specific focus on music that we didn't have so much with the theater collection. And this year, I had the great pleasure and privilege to work with the Canadian Arts Presenting Organization and its partners in cooperation that has reinforced obviously the international dimension of my work, but which had also the merits of adding another perspective on performing arts data, the perspective of the primary value chain made up of artists, of production companies, of presenting organizations and operators of arts facilities. Yet, this is just one personal journey. What has struck me most on my journey was the fact that I keep meeting other people from various countries who have a very similar vision to the one we had started to articulate in cooperation with the Swiss theater collection. The vision of an international knowledge base for the performing arts based on linked data technology. The journeys of all these people typically start from one specific area of the performing arts, of this performing arts value network with a particular set of stakeholders in mind and with a focus on a specific number of usage scenarios. So in the first phase of K-PA-CoA's Linked Data Future Initiative, we have connected these dots and systematically explored the different stakeholder views on the data infrastructure we are about to build. I won't bother you with all the different usage scenarios. You will find them in the report. We have differentiated between some 20 stakeholder groups. So we have the performing arts professionals, we have personal agents, we have casting agents, production companies, presenters, promoters, ticketing platforms, operators of arts facilities, writers and composers, theater concertgoers, online consumers, media professionals, tourist boards, persons providing information to tourists, search engines and computer-based personal assistants, educators and learners, researchers, lexicographers, heritage institutions, private collectors and statistical services. To simplify things, we have assembled these different stakeholder groups into seven larger categories, which are, you can see it on the slide here, production, presentation and promotion, coverage and reuse, live audiences, online consumption, heritage, as well as research and education. So based on the usage scenarios, as I mentioned before, we have asked, what kind of data will these different groups need? In an overview table, we have then highlighted the data that is common to all of them. And believe it or not, there's quite some overlap when it comes to the data that's relevant for the different key stakeholders. And it would only make sense for them to cooperate more closely, to leverage the many synergies that linked open data brings about. So the idea, I'm missing a slide, so the idea is to create a linked open data ecosystem consisting of several layers. We have a data layer consisting of decentralized data sources and several data hubs aggregating and interlinking all the data. We have then a semantic layer consisting of a shared data model, a so-called ontology. Then we have an application layer consisting of various cloud services for data extraction, data analysis and data visualization. And last but not least, we have a presentation layer. The actual platforms and software applications, the end users are going to interact with. Now this is the concept, but where do we stand regarding its implementation? So as we are missing a slide, I'm going to skip further and... Actually, I can retrieve it. Can we retrieve it? For those who are local, because it actually is still in there, it was just hidden. So... No, it's still this one. Oh, not this one. Okay. Sorry about that. So let's go back in slide mode. Sorry about that. So one important thing where these different parts that have been implemented so far is the common ontology. So this is a shared data model that has the function of a common language that allows the different elements of the knowledge base and eventually their users to communicate with each other. Another key aspect in view of the interoperability of the various interconnected databases are base registers, or so-called authority files. By referring to the entities listed in these base registers, we can store statements about the same entity. Let's say about a particular opera in different databases and use federated queries to retrieve this information as if it came from one single source. In fact, base registers contain the lists of named entities of our common language. In the case of the performing arts, we are, for example, interested in creative works. We're interested in their additions and character roles in performing arts venues in humans or in organizations. These base registers play a key role in interlinking datasets from various sources. Many of these registers of named entities already exist. I've provided you with some statistics for Wikidate alone, which is just one of these databases. Now let's turn to the Canadian implementation partners. RIDO and Culture Creates are the primary implementation partners in the initial phase of the LDF initiative. RIDO-SENPRO is a centralized information system for the performing arts production market. It integrates and facilitates a number of operations, such as showcase application, event registration, and block booking. The goals of SENPRO are to reduce manual data population to different systems and to enable the reuse of core data across a number of business processes. SENPRO will just gather a lot of reliable, validated, up-to-date information about performing arts companies, productions, and presenters. Such information will then populate directories and applications intended for sector stakeholders. In addition, all nonsensitive information will be fed into the Canadian knowledge graph for the performing arts. Culture Creates, a startup company from Montreal, has been developing tools and services to enhance the findability and discoverability of performing arts events in the consumption market. Its footlight tool uses natural language processing and harvests unstructured or semi-structured event information on websites and translates it into machine-readable structured metadata. Up on validation of the metadata by the presenters of these events, the metadata is republished as JSON-ALD, embedded in the organization's websites, and also fed into the Canadian knowledge graph for the performing arts. ArtsData.ca is a Canadian performing arts knowledge graph initiated in 2018 by Culture Creates with the support of the Canada Council for the Arts and several arts organizations. Today, the graph database is still in its infancy. It's eventually meant to assemble all relevant data about current and future performing arts events in Canada by Canadian artists or artists' collectives abroad. ArtsData.ca is meant to become the Canadian hub within the international open data ecosystem for the performing arts. You can find out more about these tools and services by visiting the project website. This is also a variable to find the documentation about the data model that is being used in the context of artsdata.ca and you can explore our sample data. So, this is how the sample data looks like on the website, where you can explore it in an interactive manner. And this example happens to show the performance of Iskwe in Toronto. Let's now have a look in more detail at one of the modeling samples that we have broken down for you on several slides. I guess you probably all know Hydro-Quebec's hydroelectric power company. J'aime Hydro is a touring theatre production about that company. And as you can see on this slide, it has been produced by two production companies, Porte Parole and Chant Gauche, which, as you can see on this data model, have their postal addresses, have their offices and contact points. J'aime Hydro has been written by Christine Beaulieu, who also plays the main role. Further cast members are Mathieu Douayon and Mathieu Gousselon. Beside them, the production has many contributors which are listed here at Ramaturch, for example, a stage director, an assistant stage director, a set designer and so on. Obviously, people are also interested in knowing where they can see this show. For this reason, we also provide information about its different representations. For example, the performance on the 1st of May in Saint-Yassant, or the upcoming series of performances in Montreal next month. You can see that it starts on September the 10th and ends on September the 15th, that it will take place at Yuzin City and there is a website from where you can get tickets. But before, now this you can see, here you can see on the slide the integrated view of all this data. But now before you get off to buy your tickets for this show, let me finish this. As you have seen, many things are already in place, yet there are also still many challenges ahead. We have sketched them out in our report. I won't cover them in detail here. Let me close this presentation by mentioning the five recommendations made by the board of experts that served as our advisory committee with regard to the further implementation of K-PACOA's Linked Digital Future Initiative. So recommendation one is immediate focus should be placed on populating a Canadian performing arts knowledge graph. Remember, artstata.ch is for now mainly a structure with only still very little data available. So now data about more current and future events should be ingested into this knowledge graph from made available through interoperable data systems. Recommendation number two, Wikidata is to be seen as complimentary to artstata.ca. Efforts should therefore be undertaken to contribute to its population with performing arts related data that is of relevance in the context of this Canadian knowledge graph. Number three, a data governance framework needs to be developed in cooperation with representatives from across the entire art sector in order to establish who is able to share what kind of data with whom and who will have authority over which data and information. Recommendation number four, further research is needed to better understand user requirements with regard to the adoption of linked open data practices in service offerings. Stakeholders who are now expected to make an additional effort to contribute or enhance performing arts related data should know why they're doing so and for whom. And the fifth and last recommendation is that further effort is required to develop and describe novel business models that leverage and maintain a well functioning linked open data ecosystem for the performing arts. It is essential to evaluate the long-term economic sustainability for individual contributions to the common knowledge graph by key players in the performing arts value network. Thank you for your attention and now give it over to our project director. Thank you, Beat. So these are recommendations as an art service organization whose mission is to help this sector succeed in the 21st century, our role now is to act upon those recommendations, which we will do. So in terms of immediate next steps, we'll work with our partner partners, Culture Creates, Anna Rideau, so that we can start immediately populating that our data can be in knowledge graph with actual data because a data model on its own is not worth much. It will not be very useful to Alexa, Sairi and Google. Therefore, we need data first step. But then, as I said earlier, we need to provide digital literacy to this sector. As we're doing now, we need to go out and explain why this change is necessary, why the environment has changed and why suddenly we need to provide our information as linked open data. And then we need to provide actual guidance for whenever someone says, oh yeah, I wanna do that. Well, it's not that easy. So we'll be providing one-on-one guidance to our organizations as they want to embark on a link open data journey. Of course, there remains very critical governance questions that we need to address. Those cannot be left on the side burner for too long a time, otherwise we'll have a problem. So this will be a very important priority in the next year. But then in the longer term, we have to bear in mind Canada is small. If this is to succeed, we have to bring it to the international level. We're already doing that by working with Swiss partners, but we know that some of the people are doing similar work in Belgium, in Australia and in other places in the world. So eventually we need to also bring all of these stakeholders together into an international linked open data ecosystem for the performing arts. That's the long-term goal. So before we wrap up, I'd like to direct you to the Linked Digital Future website. Well, you'll be able to find this report. You'll be able to explore the data model. But also for those who are still new to linked open data, you can get some guidance and some help to find out what actually linked data means. It's a number of things. So we won't go into a detailed explanation in there, but you'll have more resources on our website. I need, I have many, many people to acknowledge and thank. I can't do all of them today, but at least I want to thank the members of my advisory committee. Jean-Roberte Bisayon, Clément Laberge, Margaret Lam, Tammy Lee, Mariel Marshall, and Marie-Pierre Pilotte. This data model would not have come about without the help of Gregory Sommier-Finch from Culture Creates and Adrien Schvent from Zatzuko. And of course, we had many, many, many more contributors. So I invite you to take time as you read the report to go all the way to the end to see all of the wonderful folks who contributed in one way or another. Finally, I'd be remiss not to mention our funding partners. This initiative would not have been possible without the financial support of the government of Canada and of the Canada Council for the Arts. That's it. Thank you for, if we have any questions here in the room or else on Twitter, we'll be happy to take them. So I have a couple of questions that might be relevant to the art sector to maybe just pass along right now. What governance frameworks have you begun to explore in the process of beginning to look at the Canadian Knowledge Graph? Good questions. There are different levels of governance that will need to be addressed. There are ethical matters. As Béat mentioned, we need to define who will have access to the data and under which conditions. For example, let's presume that in a year from now we have massive data about the performing arts in artsdata.ca and suddenly commercial applications built off of this data become possible. Well, who should benefit from those profits? Ultimately, if the data belongs to the sector those profits need to go back to the sector which calls for certain governance structure that be it on a cooperative structure or enough for profit. We still have to define that but this is one area of governance that we need to clarify. Then there is also the technical governance, the maintenance of the data model itself. The data model is in fine shape as it is but it's still not complete. There are still questions that we have not resolved. We have not dug into what are called control vocabulary to define at a more granular level the disciplines, the types of organizations, types of places, types of audiences and so on. This needs to be done in collaboration with the sector. We will need probably another form of governance structure to address these questions and to make sure that this data model remains relevant and fully adapted to the real world in the future. So in terms of critical diversity and engaging diverse world views what steps is it linked to digital futures initiative taking in order to address these questions? This is probably the most challenging question and the one that has been the nagging thing in my brain for the longest time in the last year. We have worked really hard to codify all information and concept that pertains to the performing arts domain as we know it. The performing arts domain in and of itself may be carrying some colonial structures. It may not be reflective of all of the diversities of cultural expressions in the world. So we may begin with a data structure that is already biased. And then there are many, many concepts or elements of diversity that belong to other knowledge domains outside of the performing arts. Let's take for example, indigenous knowledge. So we could definitely link performing arts data with information about the indigenous identity of an artist and that identity itself would be linked to a First Nation or an Inuit or Metis people for example. However, it does not belong to, at least not to me as a settler to start providing linked open data about First Nations. This is a task that belong to the peoples themselves. So our role then becomes to raise awareness of the need to undertake that which we've done in the report and throughout our research and we're really into an action research mode. So we were constantly consulting and at the same time raising awareness and trying to come together to a greater level of knowledge. So ultimately we'll have to continue these discussions with indigenous stakeholders and other stakeholders that represent other aspects of diversity so that eventually they become part of that linked open, well not part of the linked open data ecosystem for the performing arts but that they are linked to it. So in terms of first steps, what does an arts organization need to have in place in terms of a basic toolkit in order to begin engaging in moving towards a linked open data and what phases of this project are in place to help them along the way? Very good questions. There are, as Bette said, many trajectories, many paths forward. So we do have a data model but there are a number of ways that data can be coded into that model and that data model and the idea of the large graph isn't the be all and all. As Bette said, buke data is one means by which we can provide linked open data to search and recommendation technologies. So on our end, we are working with first an operator of an information system for the performing arts. So this service organization has this massive system, they'll be conducting tons of data, so we'll just try to eventually import it into the knowledge graph. That's one way of doing it. Then the other way is to structure the information according to vocabulary such as schema and others so that it's made available for source engines. We're doing that with cultural creates but for people who may be watching us on HowlAround, there are other ways that you can start to structure your event metadata following the schema vocabulary. You can look out for different plugins that can help you do that. So this is another way and I'll go back to wiki data. Anyone, anyone can be an editor of wiki data. It's not rocket science. So if there happens to be a wiki data entry about your arts organization or about venues that you are using, just go and check those. Are they accurate? Are they complete? Could you add more information, more properties about these different entities as wiki data calls them? So maybe one final question and that would be why is discoverability a collective issue and not just one for individual arts organizations? That would require an entire presentation but I don't, we have time for that now. I think what's important to understand is that arts organizations are not competing with one another and arts organizations can only present so many performances per year. The object of the performing arts cannot be mass produced. You can only present one event at a time in your facility that's it and you need to set up and tear down. So in order to provide audiences with a critical mass of options to attend live performances, we need to work together. The real competition comes out in many forms in the entertainment sector. There's the movie industry, there's Netflix but people also want to go to restaurants and they can easily do that instead of going to a show. So unless we get our grips or collective grips and we make our data available for people to find shows on Google, audiences might choose other entertainment options and that's not an option for me. Thank you both and with that, this concludes the launch for the Linked Open Data Report and for the Linked Digital Futures Initiative. Thank you all for participating and for joining us today. Thank you for being there. Done.