Critical issues presentations/GLAM data on Wikidata – Taking GLAM relationships to the next level

From Wikimania 2016 • Esino Lario, Italy
Jump to navigation Jump to search
Submission no. 177
Title of the submission

GLAM data on Wikidata – Taking GLAM relationships to the next level

Author of the submission
  • André Costa
  • John Andersson
  • Axel Pettersson
Country of origin

Sweden; Sweden; Sweden


Projects, Technical

  • GLAM
  • Wikidata
  • references
  • sources
  • collaboration
  • data

This talk addresses the following issues:

  • There is a need to develop and revitalize many ongoing GLAM collaborations.
  • Many GLAMs which want to collaborate do not have content suitable for Commons/Wikisource (either due to copyright or format).
  • There are lots (49%) of unsourced statements on Wikidata (and only 21% which are not referenced to Wikipedia)[1].

The aim is to allow other organisations to replicate the actions we have taken to address these issues, thereby opening up a new type of content and a new wave of collaborations with GLAM institutions and authorities.

The Wikimedia movement has a strong track record when it comes to collaborating with various GLAM institutions. These usually take the form as one of 1) increased access for Wikimedians to the collections; 2) the involvement of subject experts in editing the Wikimedia projects; 3) batch uploads i.e. making available a large body of content for direct use by Wikimedia (normally on Commons/Wikisource). Out of the three the last is traditionally the one which requires the least work by the institution itself. As such it is often an easy start for a GLAM institution which has decided to get involved.

But for many GLAMs motivation alone is not enough since they might be limited by such things as the copyright (e.g. a museum of modern art) or the format (e.g. tactile objects, larger works etc.) of their collections or simply lack the possibility of digitising them. These GLAMs have one resource where they normally own all the rights and where their expertise really comes to light, their data. During the process of curating, analysing and researching their collections GLAMs normally build a database of not only the information directly related to their collection (samples, paintings, shoes, houses, ...) but also about their context (people, organisations, movements, geography, history, ...) and a structure for how all of this is connected (relationships, hierarchies, ...). And thanks to the newest member of the Wikimedia family there is now a home for much of this data.

This data provides an invaluable resource for Wikidata in that it allows us to enhance the data within the domains of expertise of each GLAM. It also allows us to identify missing entries and connections and most importantly it allows us to reference every single fact we add using an authoritative source. Additionally we can source any pre-existing fact which agrees with the GLAM data allowing us to actively combat the issue of unsourced statements non-sources such as references to Wikipedia.

As an example we will look at how two pilot projects allowed us to add over 10,000 references to Wikidata as well as enriching it with new facts. What are the things to keep in mind, what works, what didn’t. We’ll also look where we are heading next and how we can apply the same methodology to data from national authorities in order to incorporate and source the information about cultural heritage data.



Not accepted without fault (a prior submission from the same author(s) was accepted)