BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:A New Dataset and Method for Automatically Grading ESOL Texts - He
 len Yannakoudakis\, University of Cambridge
DTSTART:20110610T110000Z
DTEND:20110610T113000Z
UID:TALK31695@talks.cam.ac.uk
CONTACT:Thomas Lippincott
DESCRIPTION:We demonstrate how supervised discriminative machine learning 
 techniques can be used to automate the assessment of 'English as a Second 
 or Other Language' (ESOL) examination scripts. In particular\, we use rank
  preference learning to explicitly model the grade relationships between s
 cripts. A number of different features are extracted and ablation tests ar
 e used to investigate their contribution to overall performance. A compari
 son between regression and rank preference models further supports our met
 hod. Experimental results on the first publically available dataset show t
 hat our system can achieve levels of performance close to the upper bound 
 for the task\, as defined by the agreement between human examiners on the 
 same corpus. Finally\, using a set of 'outlier' texts\, we test the validi
 ty of our model and identify cases where the model's scores diverge from t
 hat of a human examiner.\n
LOCATION:FW26\, Computer Laboratory
END:VEVENT
END:VCALENDAR