BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Modeling coherence in ESOL learner texts - Helen Yannakoudakis\, U
 niversity of Cambridge
DTSTART:20120525T113000Z
DTEND:20120525T120000Z
UID:TALK38344@talks.cam.ac.uk
CONTACT:Ekaterina Kochmar
DESCRIPTION:To date\, few attempts have been made to develop new methods a
 nd validate\nexisting ones for automatic evaluation of discourse coherence
  in the\nnoisy domain of learner texts. We present the first systematic an
 alysis\nof several methods for assessing coherence under the framework of\
 nautomated assessment (AA) of learner free-text responses. We examine the\
 npredictive power of different coherence models by measuring the effect\no
 n performance when combined with an AA system that achieves competitive\nr
 esults\, but does not use discourse coherence features\, which are also\ns
 trong indicators of a learner's level of attainment. Additionally\, we\nid
 entify new techniques that outperform previously developed ones and\nimpro
 ve on the best published result for AA on a publically-available\ndataset 
 of English learner free-text examination scripts.
LOCATION:FW26\, Computer Laboratory
END:VEVENT
END:VCALENDAR