Dear Wikidata team,
I am writing on behalf of the YAGO team at the Max Planck Institute for Informatics in Saarbruecken [1]. We have heard about the Wikidata project, and we are very excited to learn that you aim to launch a free knowledge base in the spirit of Wikipedia.
We would like to get in touch with you -- also to see whether or how we could help on the long run. Let me briefly tell you what we have on our side: As you might know, YAGO is knowledge graph that has been extracted automatically from the infoboxes and categories of Wikipedia. We have evaluated YAGO manually and achieved a precision of 95%, meaning that statistically speaking, only 5 out of 100 statements in the knowledge graph are extracted wrongly. We also have a link of the Wikicategories to the WordNet taxonomy (again with 95% precision), and type checking methods for the extracted statements. Should these things ever be useful to you, we would be happy to help.
I will be at the WWW conference next week. In case some of you are there, too, I'd be happy to get in touch to learn more about your current work.
Thanks
Fabian
On Fri, Apr 13, 2012 at 3:50 PM, Fabian M. Suchanek f.m.suchanek@gmail.com wrote:
Dear Wikidata team,
I am writing on behalf of the YAGO team at the Max Planck Institute for Informatics in Saarbruecken [1]. We have heard about the Wikidata project, and we are very excited to learn that you aim to launch a free knowledge base in the spirit of Wikipedia.
We would like to get in touch with you -- also to see whether or how we could help on the long run. Let me briefly tell you what we have on our side: As you might know, YAGO is knowledge graph that has been extracted automatically from the infoboxes and categories of Wikipedia. We have evaluated YAGO manually and achieved a precision of 95%, meaning that statistically speaking, only 5 out of 100 statements in the knowledge graph are extracted wrongly. We also have a link of the Wikicategories to the WordNet taxonomy (again with 95% precision), and type checking methods for the extracted statements. Should these things ever be useful to you, we would be happy to help.
Great to hear.
I will be at the WWW conference next week. In case some of you are there, too, I'd be happy to get in touch to learn more about your current work.
Denny Vrandečić will be there for Wikidata.
Cheers Lydia