About the

Yago Ontology

Yago (”yet another great ontology”) is the baseline source for named entity instances used in UMBEL.

These instances total about 1.5 million notable things and topics derived from Wikipedia.  UMBEL uses Yago instances because of the strong typing of their relationships and high quality.  Importantly, Yago is also used because its mapping to the WordNet linguistic database provides a ready pathway for relating to UMBEL subject concepts.

This online demo is basically a Web service retrieval point with limited functionality for these Yago instances.  Nonetheless, it does represent the first time these Yago instances have been exposed as linkable data.  Yago instance data relationships are viewed through the Zitgist DataViewer (though the URIs can be supplied to any RDF browser).

Yago was developed by Fabian M. Suchanek, Gjergji Kasneci and Gerhard Weikum at the Max-Plack-Institute for Computer Science at Saarbrücken University. A WWW 2007 paper and more recent and detailed technical report describing the project is available, plus there is an online query point for the Yago service and the data can be downloaded.

Yago's Wikipedia and WordNet combination provides a rich hierarchical ontology with high precision and recall. The system contains nearly 100 data relations (predicates).