Here's your quick overview of what has been happening around Wikidata over the last week.
- Open request for adminship: Wiki13
- Tool of the week
- SQID allows you to analyse, browse and query Wikidata. SQID is inspired by Magnus Manske's Reasonator, but focuses on prominently featuring information about Wikidata classes and properties.
- Did you know?
- Newest properties:
- General datatypes: height of center of mass, road number formatter, Vietnamese middle name, heraldic attitude, traffic sign template image
- External identifiers: Tree of Public Interest ID, Denkmaldatenbank Thurgau ID, DSSTOX compound identifier, South Africa EMIS code, Archive Site Trinity College Cambridge ID, WISAARD resource ID, Gateway to Research organisation ID, SÚKL code, Science Fiction Awards Database author ID, Power plant operating licence (Turkey)
- New property proposals to review:
- General datatypes: certified as, number of stages, convergence rate, step count, Alternative form, view, version type for works, advertisement copy
- External identifiers: LibraryThing series identifier, Swiss Industrial Heritage ID, TOPCMB ID, Library of Congress Medium of Performance Thesaurus ID, Sports-Reference.com college basketball school ID, SAT-matrikulo, Signal number, BHF author ID, BHF magazine ID, SPLC Group ID, SPLC Individual ID, Open Civic Data Division Identifiers, RKD thesaurus ID, TCLF ID, Presence compositrices ID of composer, Presence compositrices ID of work
- Query examples:
- Commons queries:
tinyurl.com/y25qa8ef (source, at 53:45)
- Animal hugs:
- The last week was our quarterly prototyping week. We worked on the following projects. None of them are ready for prime-time yet but we'll continue with them.
- Slices: We've had a lot of requests for accessing dumps of a smaller part of Wikidata's data since rarely anyone needs the complete data in Wikidata. The tricky part is figuring out which part is needed and if any of that can be generalized. We looked into for example how to make dump generation faster so we could potentially produce more smaller dumps that only cover a part of Wikidata's data, either thematically (e.g. humans) or by type of data (e.g. only statements and English labels and aliases but not sitelinks or descriptions).
- REST API: As part of our effort to make it easier to access Wikidata's data for programmers we looked into a REST API. We tried to see if we could cover the existing action API modules in a REST API. We could. We'll take this as input for our ongoing API work now.
- Improving quality ratings through ORES: ORES can judge the quality of an Item automatically. It is currently not very good at it however. We tried a few things to make it more accurate and found some easy wins we'll probably make happen in the next weeks.
- Query manipulator: One of the ways we could potentially improve the load situation of the Wikidata Query Service is by automatically analyzing and then redirecting a bunch of queries to other systems that are more suitable for that particular type of query. The nice thing about that would be that the person/program sending the query wouldn't have to care about it but it'd be done automagically for them. We tried to build such a system and the results look very promising but more work/experimenting is needed, especially together with the WMF Search team.
You can see all open tickets related to Wikidata here. If you want to help, you can also have a look at the tasks needing a volunteer.