BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:A Seed-driven Bottom-up Machine Learning Framework for Extracting 
 Relations of Various Complexity - Diarmuid Ó Séaghdha (Computer Laborato
 ry)
DTSTART:20090518T120000Z
DTEND:20090518T130000Z
UID:TALK18358@talks.cam.ac.uk
CONTACT:Diarmuid Ó Séaghdha
DESCRIPTION:At this session of the NLIP Reading Group we'll be discussing 
 the following paper:\n\nFeiyu Xu\,Hans Uszkoreit and Hong Li. 2007. "\nA S
 eed-driven Bottom-up Machine Learning Framework for Extracting Relations o
 f Various Complexity":http://www.aclweb.org/anthology-new/P/P07/P07-1074.p
 df. In Proceedings of ACL-07.\n\n*Abstract:*\nA minimally supervised machi
 ne learning framework is described for extracting relations of various com
 plexity. Bootstrapping starts from a small set of n-ary relation instances
  as “seeds”\, in order to automatically learn pattern rules from parse
 d data\,\nwhich then can extract new instances of the relation and its pro
 jections. We propose a novel rule representation enabling the composition 
 of n-ary relation rules on top of the rules for projections of the relatio
 n. The compositional approach to rule construction is supported by a botto
 m-up pattern\nextraction method. In comparison to other automatic approach
 es\, our rules cannot only localize relation arguments but also assign the
 ir exact target argument\nroles. The method is evaluated in two tasks: the
  extraction of Nobel Prize awards and management succession events. Perfor
 mance for the new Nobel Prize task is strong. For the management successio
 n\ntask the results compare favorably with those of existing pattern acqui
 sition approaches.
LOCATION:GS15\, Computer Laboratory
END:VEVENT
END:VCALENDAR