Wikidata:WikidataCon 2017/Notes/Structured Data on Wikimedia Commons: what's coming, and how to be involved as Wikidatans

Title: Structured Data on Wikimedia Commons: what's coming, and how to be involved as Wikidatans

Speaker(s) edit

Name or username: Sandra Fauconnier User:Spinster

Contact (email, Twitter, etc.): sfauconnier wikimedia.org / https://twitter.com/sanseveria

Abstract edit

The Wikimedia Foundation and WMDE are currently working to integrate structured data support with the the software that enables Wikidata, Wikibase, into Wikimedia Commons. Structured data will turn Wikimedia Commons into a more engage-able platform from editors and readers, improving discovery, organization and the pathway for small contributions in all of the languages supported by Wikimedia projects. Additionally, providing a better Metadata infrastructure allows for better uploading of content from partner communities who want to share their media and to dynamically take advantage of the Wikimedia community's environment. Like Wikidata, structured data on multimedia files will allow for more dynamic results and improvements to how Commons content can be used both within and beyond the Wikimedia movement.

Sandra Fauconnier, community liaison, will explain the updated plans for the project, with a focus on technological developments, and what we are learning from community engagement and research.

Together, the attendees of this session will identify the challenges that we expect to be upcoming, and discuss ways in which the Wikidata community can be involved in the upcoming years.

Collaborative notes of the session edit

Wikimedia Commons: https://commons.wikimedia.org

The images come from:

  • Own work by Commons contributors
  • Found under free license on Ze Internetz
  • Files from partner institutions

There are social systems for highlighting great content : https://commons.wikimedia.org/wiki/COM:FP

Example Video and file page

Ten days that shook the world, the Russian October 1917 revolution

https://commons.wikimedia.org/wiki/File:October_Ten_Days_That_Shook_the_World_(1928).webm

Wikimedia Commons built on MediaWiki, so uses wikitext

So there are a lot of templates

Such as the licenses template on the "Ten days that shook the world" video

Categories are the best thing we have to organize the media at the moment

Crowdsourced semi-hierarcical tagging system

Example category name that is 4 lines long

Commons can be difficult to navigate if you are not familiar with it

Search is not multi-lingual, and is very often confusing, even when searching in English

Structured Data on Commons will enable structured and machine readable metadata on Commons so search, editing, uploading is easier

Program will add structured data to Commons, not take away any of the existing categories or templates

Example of visualisation mixing Commons and Wikidata: Crotos http://zone47.com/crotos/

Identify MediaInfo Entities

User interface

Copyright and licencing CC0 versus CC-BY-SA

Overview of the session edit

Data modelling

New properties

Wikidata loves Commons

Phabricator