BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Improving Grammaticality in Statistical Sentence Generation - Yue 
 Zhang
DTSTART:20100524T113000Z
DTEND:20100524T123000Z
UID:TALK24925@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\nStephen Wan\, Mark Dras\, Robert Dale and Cécil
 e Paris. 2009. "Improving Grammaticality in Statistical Sentence Generatio
 n: Introducing a Dependency Spanning Tree Algorithm with an Argument Satis
 faction Model":http://aclweb.org/anthology-new/E/E09/E09-1097.pdf. Proceed
 ings of EACL-09.\n\n*Abstract:*\nAbstract-like text summarisation requires
  a means of producing novel summary sentences. In order to improve the gra
 mmaticality of the generated sentence\, we model a global (sentence) level
  syntactic structure. We couch statistical sentence generation\nas a spann
 ing tree problem in order to search for the best dependency tree spanning 
 a set of chosen words. We also introduce a new search algorithm for this t
 ask\nthat models argument satisfaction to improve the linguistic validity 
 of the generated tree. We treat the allocation of modifiers to heads as a 
 weighted bipartite graph matching (or assignment) problem\, a well studied
  problem in graph theory. Using BLEU to measure performance on a string re
 generation task\, we found an improvement\, illustrating the benefit of th
 e spanning tree approach armed with an argument\nsatisfaction model.
LOCATION:GS15\, Computer Laboratory
END:VEVENT
END:VCALENDAR