BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Graph-Based Methods for Large-Scale Multilingual Knowledge Integra
 tion - Gerard de Melo\, Max Planck Institute for Informatics
DTSTART:20110722T110000Z
DTEND:20110722T120000Z
UID:TALK32130@talks.cam.ac.uk
CONTACT:Thomas Lippincott
DESCRIPTION:An increasing number of applications are making use of informa
 tion about\nsemantic relationships between words\, names\, and entities or
  concepts.\nThis talk presents three graph-based methods to obtain and int
 egrate such\nlexical semantic knowledge. The first involves learning model
 s to\ndisambiguate word meanings. The second reconciles equivalence and\nd
 istinctness information about entities from multiple sources. The third\nm
 ethod adds a comprehensive taxonomic hierarchy\, reflecting how different\
 nentities relate to each other. Together\, they can be used to produce a\n
 large-scale multilingual knowledge base semantically describing over 5\nmi
 llion entities and over 16 million natural language words and names in\nmo
 re than 200 different languages. See\nhttp://www.mpi-inf.mpg.de/yago-naga/
 uwn/ for more information.
LOCATION:FW26\, Computer Laboratory
END:VEVENT
END:VCALENDAR