OPEN DATA—In our last chapter on the different levers of discoverability, we briefly examined the concept of linked data. For this third installment, we’ll take a closer look at open data. The concept is gaining ground within government and academic circles as new demands are being made in terms of transparency (see Government of Canada Open Data portal) and accessibility to publicly funded research results. So what does this mean for artist-run centres and what can it help us accomplish?
February 20, 2020
By Isabelle L’Heureux, Digital cultural development officer, Conseil québécois des arts médiatiques (CQAM), Regroupement des arts interdisciplinaires du Québec (RAIQ), and Le Regroupement des centres d’artistes autogérés du Québec (RCAAQ)
Defining the concept of open data
Open data is structured, machine-readable, royalty-free data that is accessible and reusable. This data can be statistical or geomatic, or it may correspond to coordinates, lists, plans, items in a collection, research results, and more.
Open data has a few defining characteristics. It must be accessible, that is, it must be available to all, free of charge, in its entirety, and ideally in an open, standard, editable format. The CSV format (which stands for comma-separated value), for example, meets this criteria and can be used to facilitate access to data sets. Open data must also be published under conditions that allow its re-use and facilitate cross-referencing with other data sets. Finally, open data favours universal participation, which means that its use is unrestricted, even for commercial purposes.
Within a cultural context, many kinds of open data sets are imaginable. Some organizations will publish data relating to events they have organized (list of festival concerts, attendance statistics, etc.) or to items that are part of their collections (tables or graphs—we’ll come back to the notion of graphs in the next section—with titles, names of artists, dates, materials, etc.).
It’s interesting to note that open data sets can exist at different levels within a fairly broad spectrum of openness. Not all initiatives follow the same degree of complexity. Tim Berners-Lee, the inventor of the Web, schematized the different levels of openness that internet data can have in his 5-star Open Data plan. The first level corresponds to minimal openness and is often very easy to publish, while the fifth and final level involves a maximum level of openness, and is perhaps more difficult to achieve.
🌟The first level is when a document is published on the Web under an open license. For example, a text in PDF format published under a Creative Commons Attribution licence.
🌟🌟The second level involves publishing a structured document on the Web, for instance an Excel spreadsheet.
🌟🌟🌟The third level involves publishing structured information in an open and non-proprietary format such as CSV.
🌟🌟🌟🌟At the fourth level, each element in the data set (object, person, relation) is identified by a URI, or Uniform Resource Identifier, which gives it an unambiguous, perennial identity that can be referenced the same way anywhere on the Web. For example, a publication’s ISBN can play the role of a URI on the Web.
🌟🌟🌟🌟🌟The fifth level is achieved when open data is linked to other open data, which then turns it into linked open data. Several concepts are associated with this level: linked data, Semantic Web (and Web 3.0 for others).
The advantages of open data
For communities and individuals, there are many benefits to opening up their data.
- for government entities, having open data can indicate a willingness to maintain transparency;
- it supports innovation by providing access and opportunities to reuse information and knowledge;
- greater availability also benefits the world of research;
- in some contexts, it can help people make more informed decisions.
The Canada Council for the Arts gives access to their data tables on grant recipients, which gives the art community a statistical portrait of arts funding across the country. Based on these, organisations can reflect upon, investigate, identify trends, and position themselves within this overall profile and have a more informed opinion. Opening data can mean opening a dialogue. As a practice, it can be embedded in values of engagement, sharing, collaboration, and the creation and circulation of knowledge.
As seen earlier, open data can be modified in relation to the context or technical expertise at hand. Of course, not every Canadian cultural organization has a Semantic Web specialist on staff. The cultural milieu is composed of passionate, competent professionals who, for the most part, work on multiple fronts simultaneously often with very limited resources. Therefore, it may not be relevant for everyone to invest in complex open data projects. Nevertheless, it seems useful to remember that some open data projects can be part of an organization’s mission or values, they can be quick and simple to implement, and have positive and surprising impacts. For example, any organization that has a website can publish data, whether in text or PDF, HTML, CVS or other format (level 1, 2, or 3 in the Berners-Lee Open Data Plan) under a free licence. Other initiatives rely on existing infrastructure, such as Wikidata, or open data portals (i.e. Données Québec).
Technical expertise is most required for projects that involve the attribution of URIs and linking between several data sets (levels 4 and 5 of the Berners-Lee Open Data Plan). It’s better to call on relevant experts, often from outside of the organization, to accurately assess what resources (time, money, staff) are necessary and to create the proper conditions to successfully complete the project.
In many cases, it’s good to anticipate the need for updating and versioning. Will the data set remain unchanged or will it be revised annually? Will it be replaced by a new data set after a pre-determined period? Different answers and processes will apply depending on the type of data and the publishing organization’s objectives.
Legal issues should also be considered in open data initiatives. It is vitally important to ensure that the data to be published under free licence is not subject to any prior restrictions on use or distribution. It’s therefore preferable for an organization to own the data it wants to open, or for the data to be understood as facts (gallery addresses, titles of works, etc.) and thus not protected under Canadian copyright. To guarantee that no laws will be violated in the publishing of a data set, it’s good to seek legal advice from competent professionals.
For its project called Savoirs communs du cinéma, the Cinémathèque Québécoise hired Olivier Charbonneau, a research librarian with a doctoral degree in law, to prepare a preliminary report on issues of copyright in the dissemination of cultural metadata (available in French only: Enjeux en droit d’auteur de la diffusion ouverte de métadonnées culturelles). Although this document responds to a specific context, it can be useful for many Canadian cultural institutions.
Course of action
The concept of open data gives us new ways to give greater value to the content and information produced by our organizations. However, it’s important to assess what our objectives are and what costs and skills are required to make it happen. Is it simply a matter of exporting information from an existing database in CVS format and making it accessible on the organization’s website, or does the project require data to be compiled by a human or system? If the project requires original creative work, how will contributors be paid? Is making data open consistent with the organization’s mandate, finances, and other resources?
As an example, the RCAAQ will participate in open data by sharing its members’ coordinates (which it draws from ARCA directory database), on the Données Québec portal. This requires a minimal amount of time and effort and is part of the RCAAQ membership’s promotional activities. By publishing artist-run centres’ contact information in an open and structured manner, the number of access points to this type of information is multiplied and its reuse is facilitated. While this data set may not be used immediately or widely, its availability helps us consider different ways of integrating artist-run centres into a linked data ecosystem.
By reflecting on the theme of open data, we might also consider free licenses for textual content produced by the centre (newsletters, publications). In the right circumstances (organization’s values, intellectual property, business model), these publications can be made freely accessible in a non-proprietary format (PDF, HTML) on the centre’s website or through the e-artexte digital repository. Felicity Tayler thoroughly and critically explores this idea in her essay “Situating Artist-Run Publishing Within Digital Culture” in the Grey Guide to Artist-Run Publishing and Circulation, published in 2017 by ARCA.
Through the concept of open data, we might also find a way to add value to an archive by extracting certain data to facilitate access to it (i.e., present a centre’s entire list of exhibitions and the curators and artists involved; analyze the evolution of an organization’s membership through the years). Grunt Gallery’s Activating the Archive is an ambitious and inspiring project that certainly reflects the logic of open data.
Finally, for more ambitious projects, it can be helpful to associate your organization with a university partner. Precious alliances can be found in departments such as Information Sciences, Art History, Cultural Studies, and with researchers in the field of Digital Humanities.
We hope this chapter has strengthened your collective grasp on the concept of open data, and that it will inspire relevant and daring initiatives across the Canadian artist-run centre community. In our next chapter, we will continue to explore new organizational methods and access to information in the digital realm through the enigmatic concept of the knowledge graph.
—> Next brief : Cloud Computing [upcoming link].