BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Automatic speech act identification in business emails - Rachele d
 e Felice - University of Nottingham
DTSTART:20110218T120000Z
DTEND:20110218T130000Z
UID:TALK29906@talks.cam.ac.uk
CONTACT:Thomas Lippincott
DESCRIPTION:This talk presents an approach to automated speech act recogni
 tion in business English emails. A maximum entropy classifier is trained o
 n a combination of lexical and syntactic features\, achieving 79% accuracy
  on non-native English data. I also discuss how varying conditions such as
  feature selection and training data affect performance.\n
LOCATION:FW26\, Computer Laboratory
END:VEVENT
END:VCALENDAR