Sunday, December 10, 2017

When #Wikidata is good for something

When #Wikidata is good for something, it shines. It does not take much prodding to find people to improve on what it does so well and consequently when Wikidata is useful, quality follows easily.

The promise of  a useful Wikidata was delivered at its start by having it replace the native interwiki links of Wikipedia. Within a month the quality of Wikipedia links had improved dramatically and at this time corner cases are still worked improving quality even more.

The WikiCite project is really important in many respects and it has so much more to offer. It is useful because it brings many initiatives and projects together under one roof. It is why scientific papers are included, including its authors. We find that more and more authors are included as well and they are often linked to the ORCID, VIAF and other external identifiers of this world. This has great value because it allows Wikipedia articles and information maintained elsewhere to be linked. What it can be used for is limitless. End users will find new and interesting ways to use the data and make it into information.

When Wikidata is to be good for Wikimedia projects, this information brought to Wikidata because of WikiCite has great potential. It largely reflects the citations in all the Wikipedias and consequently through linked so external sources we could know what sources are problematic, retracted or bought by interested parties. We could, we don't. When we did, we would provide weight against propaganda and fake news.

The big thing holding us back is trust. Wikipedians need to consider a Wikidata that is not only used for links and that can be trusted for high level maintenance of its citations. Wikidata is to appreciate its use and trust that its information will be used and that this will increase its value and quality. WikiCiters have to understand that Wikidata is not a stamp collection only including publication data. It must include information about retractions, about papers considered problematic for political or scientific reasons (or both).

When Wikidata is to be good for something; we should expand our collaboration with Cochrane, Retraction Watch and organisations like it. There is everything to gain; quality, contributors and relevance.

No comments: