BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Generalizing Dependency Features for Opinion Mining - Awais Athar 
 (University of Cambridge)
DTSTART:20100301T123000Z
DTEND:20100301T133000Z
UID:TALK23556@talks.cam.ac.uk
CONTACT:Diarmuid Ó Séaghdha
DESCRIPTION:At this session of the NLIP Reading Group we’ll be discussin
 g the following paper:\n\nMahesh Joshi and Carolyn Penstein-Rosé. 2009. "
 Generalizing Dependency Features for Opinion Mining":http://www.aclweb.org
 /anthology/P/P09/P09-2079.pdf. In Proceedings of ACL-IJCNLP-09.\n\n*Abstra
 ct:*\nWe explore how features based on syntactic dependency relations can 
 be utilized to improve performance on opinion mining. Using a transformati
 on of dependency relation triples\, we convert them into "composite\nback-
 off features" that generalize better than the regular lexicalized dependen
 cy relation features. Experiments comparing our approach with several othe
 r approaches that generalize dependency features or ngrams demonstrate the
  utility of composite back-off features.\n\nAs this is a shorter-than-usua
 l paper\, the presentation will also draw on the following paper as backgr
 ound:\n\nShotaro Matsumoto\, Hiroya Takamura and Manabu Okumura. 2005. "Se
 ntiment Classification Using Word Sub-sequences and Dependency Sub-trees":
 http://www.springerlink.com/index/n9ep5q1gcupeya8r.pdf. In Proceedings of 
 PAKDD-05.
LOCATION:GS15\, Computer Laboratory
END:VEVENT
END:VCALENDAR