Improving an academic organisation’s presence on the web through semantic data

CERN, the European Organisation for Nuclear Research, located in Geneva, Switzerland, was founded in 1954. Even before its first experiments started operating, the Organisation was very clear on wanting to be as open as possible with the data and research it would be producing. In 1993, CERN became a pioneer in the field of open source and Open Access by putting the software behind the World Wide Web into the Public Domain. Today, it is fighting on multiple fronts for these ideals, whether developing new software that will help share data, analyses and research openly, or being the driving force behind historic agreements, between publishers of High- Energy Physics (HEP) literature and institutions who publish in that field, aiming to make every single HEP article Open Access.The internet, and more specifically the Web, is the perfect substratefor open and collaborative projects. For the last 20 years, one method for this openness has become increasingly ubiquitous: the semantic Web. An idea launched by the inventor of the Web himself –SirTim Berners-Lee – togetherwith James Hendler and Ora Lassila, its intention was to go beyond the original principle of the first Web (Web 1.0) that had human-readable documents (Web pages) on separate servers, whicha human had to “visit” to read. The semantic Web adds semantics (here, machine-understandable meaning) to data, and makes it findable and interlinkable. This Bachelor project posits that collaborative Web projects such asthe ones under the Wikimedia Foundation umbrella, and more specifically Wikidata whichusessemantic technologies, are the perfect medium to help CERN be even more open and discoverable on the Web. It starts by looking for data about CERN that is interesting and harvestable (from INSPIRE –a HEP repository), then goes on to test pushing that data to Wikidata, and finishes by trying to assess the effect of the changes. Thisassessment being close to impossible to carry out, thethesis concludes with a few aspects that show how this kind of work canneverthelesshave a great deal of value.The two deliverables of this project were programs harvesting data from Inspire and transforming it into batches of semantic statements ready to be uploaded to Wikidata. The first, dealing with links between CERN people and their publications, created close to 6300 statements about 270 authors. The second, dealing with links between people and CERN, created 68 statements, representing an increase of over 12% of suche links already on Wikidata