Wikidata:WikidataCon 2017/Notes/Employing Wikidata for Fostering Scholarly Research

Abstract

Wikidata follows a formal and data-oriented approach to the representation of information and assertions which contain such information. This is challenging especially for non-technical experts from the humanities, where the discursive function of text is a constitutive feature of information. These disciplines not only publish but also work on the basis of texts. Consequently, the usage of Wikidata in these disciplines needs mediation and support, to bridge the different thought systems and allow scholar to integrate existing research results into Wikidata easily.

Building on a use-case we will present (a demo which shows) how a workflow for such a mediation as well as the underlying tool chain can look like. The use-case is embedded in the Film and Media Studies Open Access journal Apparatus. We show an annotation tool (neonion) which integrates with the articles of the Apparatus journals and which allows to create triples in a non-technical way by annotating relevant parts of the text. The tool is tightly coupled both with the recommender software Snoopy and Wikidata. Snoopy supports the annotation process by recommending suitable and missing properties from Wikidata for specific items. We aim to automatically ingest the annotations into Wikidata (at the moment, test.wikidata.org is used) without further effort by the creator. We believe that both workflow and toolchain are transferable to other use-cases.

The presented work shows how Wikidata’s knowledge base can benefit from scholarly research results and how scholars can contribute to Wikidata without the need to understand its logic completely. Furthermore, it demonstrates how the consistent use of properties in Wikidata can be supported and how the creation of data into Wikidata can be tied more closely with research contexts