Wikidata:WikiProject Freebase

WikiProject Freebase exists to coordinate the migration and use of data at Freebase. Freebase was a community curated knowledge base like Wikidata. It was managed by Google who in December 2014 said that they will close Freebase and offer its content to Wikidata. See also the Wikidata Freebase FAQ.

An initial import was attempted in 2015, working through over 10% of the Freebase collection. The rest remains to be processed as of 2018. For the latest, see the talk page.

Background

• On 16 December 2014, Freebase announced that it was closing and that it wished to give its resources to Wikidata.
• On 31 March 2015, Freebase became read-only
• On 1 April 2015, the initial primary sources tools have been released, and some first data has been released
• In Q2 2015, a new KG-based Google API will be launched
• Earliest three months later, the Freebase website will close (planned for Q3 2015)
• From June to August 2015, Tpt is working from inside Google in order to map and release as much data from Freebase as possible to the Wikidata community
• Since August 2016, the main freebase-URL is redirected to google-developers. Other freebase-links in wikidata do not work any more.
• In 2016, a paper on the migration was published, From Freebase to Wikidata: The Great Migration.

Denny can speak for himself, but he has special insight into the migration

• Denny is a long-time Wikimedia contributor now working for Google
• He worked for Wikimedia Deutschland and was the Wikidata director until September 2013
• He has raised the issue of this migration in internal Wikimedia discussion channels

Current work

Tpt is currently working on a mapping tool in order to build Wikidata statement from Freebase content:

• For the mapping between Freebase topics and Wikidata items it uses a merge of a mapping created by Samsung (available on GitHub) and the mapping currently contained by Wikidata, the later being preferred for conflicting cases (6,000 cases). This merge provides 4.4M relations. The mappings not already in Wikidata are availlable in this dump.
• For the mapping between Freebase and Wikidata properties it use the collaboratively created mapping available in this subpage. Feel free to improve it.
• It adds to created statements sources from the Knowledge Vault Google project.
• Primary Sources source code is available on GitHub and welcomes external contributions.
• It works on the latest Freebase dump with some restrictions because of legal requirements (mostly long textual description, images and everything related to fictional universes) and is run time to time by Tpt.

How to help

• Suggest new datasets to import using a bot
• Help to map Freebase types to Wikidata classes
• Help to map most used Freebase topics to Wikidata items
• Help to map Freebase properties to Wikidata ones
• Add Freebase data to Wikidata using Primary Sources
• Import Freebase identifiers on existing items (although it would be better to generate a current mapping rather than using the 2 year old data set)

Overview

(71,611,020)   human: 6,376,879 (8.9%)   taxon: 2,726,046 (3.8%)   administrative territorial entity: 1,943,285 (2.7%)   architectural structure: 3,159,472 (4.4%)   occurrence: 3,898,674 (5.4%)   chemical compound: 1,188,724 (1.7%)   film: 294,370 (0.4%)   thoroughfare: 630,794 (0.9%)   astronomical object: 4,601,733 (6.4%)   Wikimedia list article: 404,454 (0.6%)   Wikimedia disambiguation page: 1,358,230 (1.9%)   Wikinews article: 195,900 (0.3%)   scholarly article: 22,574,314 (31.5%)   other P31/P279: 18,284,676 (25.5%)   no P31/P279: 3,973,469 (5.5%) Module:Statistical data/by project/classes, 2020-02-16

(1,155,127)   human: 294,823 (25.5%)   taxon: 113,069 (9.8%)   administrative territorial entity: 212,100 (18.4%)   architectural structure: 38,564 (3.3%)   occurrence: 64,001 (5.5%)   chemical compound: 8,604 (0.7%)   film: 26,974 (2.3%)   thoroughfare: 6,046 (0.5%)   astronomical object: 21,650 (1.9%)   Wikimedia list article: 1,133 (0.1%)   Wikimedia disambiguation page: 870 (0.1%)   artificial object: 111,184 (9.6%)   Wikinews article: 48 (0.0%)   other P31/P279: 118,907 (10.3%)   no P31/P279: 137,154 (11.9%) Module:Statistical data/by project/classes, 2015-07-17

(5,045,856)   human: 1,325,136 (26.3%)   taxon: 315,294 (6.2%)   administrative territorial entity: 563,681 (11.2%)   architectural structure: 237,432 (4.7%)   occurrence: 201,216 (4.0%)   chemical compound: 19,682 (0.4%)   film: 119,346 (2.4%)   thoroughfare: 37,541 (0.7%)   astronomical object: 26,683 (0.5%)   Wikimedia list article: 105,702 (2.1%)   Wikimedia disambiguation page: 302,981 (6.0%)   artificial object: 434,920 (8.6%)   other P31/P279: 521,992 (10.3%)   no P31/P279: 834,250 (16.5%) Module:Statistical data/by project/classes, 2016-01-10

Tasks

• Test and improve the primary sources tools
• More data on Freebase ID (P646), especially for properties
• Development of improved Wikidata API and possible absorbing of Freebase Query Editor. API Progress tracked here

Known issues

See also issues of Primary Sources in GitHub.

• Quality of suggested sources is not very high.
• Some bugs remains in the Primary Source user interface (the loading animation remains even if the statement is well saved...).
• New statements may come with no references at all.
• ...

Participants

1. Sjoerd de Bruin (talk) 15:21, 17 December 2014 (UTC)
2. Denny (talk) 18:07, 17 December 2014 (UTC)
3. Jura (talk) 16:44, 18 December 2014 (UTC)
4. Pasleim (talk) 16:47, 18 December 2014 (UTC)
5. Harmonia Amanda (talk) 17:02, 18 December 2014 (UTC)
6. Blue Rasberry (talk) 19:06, 18 December 2014 (UTC)
7. Daniel Mietchen (talk) 22:53, 21 December 2014 (UTC)
8. Ajraddatz (talk) 17:47, 25 December 2014 (UTC)
9. Waldir (talk) 17:23, 4 January 2015 (UTC)
10. DarTar (talk) 16:08, 7 February 2015 (UTC)
11. Tpt (talk) 17:28, 18 March 2015 (UTC)
12. Spencerk (talk) 20:43, 27 March 2015 (UTC)
13. Thadguidry (talk) 20:43, 01 January 2015 (UTC)
14. RaymondYee (talk) 16:46, 30 May 2015 (UTC)
15. Sj (talk) 12:27, 20 July 2018 (UTC)
16. ElanHR (talk) 10:02, 17 March 2019 (UTC)