Unlocking the human potential of Wikidata
Wikidata is amazing. Thanks to the amazing Wikidata evangelists out
there, I feel confident that, given at least five minutes, I can
convince anyone that Wikidata offers a critical service necessary for
Wikipedia, other wiki projects, and generally the future of knowledge. Challengers welcome. :)
But Wikidata has a problem. Right now it’s optimized to ingest and grow. We’ve written about how it’s not ideal for maintenance of the datasets, but automated ingestion of datasets is what Wikidata does best.
All this automated growth doesn’t necessarily connect well with the organic growth of Wikipedia and other projects. And we can see that Wikidata hasn’t truly captured the positive attentions of existing editor communities.
For that human touch Wikidata needs ever so much, it must reach out to the projects that gave rise to it.
One idea for doing this would be to make human editing of Wikidata
easier. Make editing Wikidata as easy as adding a citation to
Wikipedia. Literally.
Highlight a statement, click a button like “Structure this statement” and grow Wikidata, all without leaving your home wiki.
What it might look like to edit Wikidata from Wikipedia.
While Wikitext will always have its place for me, I’ve quite warmed up to the visual editor, and prefer its interface for adding citations. While it’s only an idea for an experiment at the moment, how might an inline Wikidata editor following the same pattern could be change the game for Wikidata?
A lot of data is already citing back to home wikis, but a powerful enough editor could pull the citation on a statement through to the Wikidata entry, along with the import source.
So much data is already coming from Wikipedia, but humans can do even better.
I have always thought it would be great to see Wikidata’s support for multi-valued properties leveraged more fully. A language-agnostic knowledgebase will be a new space to compare and resolve facts. A meeting of the many minds across different languages of Wikipedia could spell better information for all.
An advanced enough system could encourage contributions on the basis of coverage, highlighting cited statements which have not yet been structured.
And of course, at the very least, we speed up building an intermediary representation of knowledge, not tied to a specific language. People sharing knowledge across wikis, helping to further bootstrap a Wikidata community with close ties to its older siblings.
granolatech reblogged this from superdrivel
superdrivel reblogged this from hat-note
ventolinmono liked this
peopledon-tknow liked this
hat-note posted this