BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:A Bayesian Approach to Learning the Structure of Human Languages -
  Phil Blunsom\, Department of Computer Science\, Oxford University
DTSTART:20120530T131500Z
DTEND:20120530T141500Z
UID:TALK37845@talks.cam.ac.uk
CONTACT:Stephen Clark
DESCRIPTION:Grammar Induction has long been a central challenge of Computa
 tional\nLinguistics. Empirically demonstrating the ability of computationa
 l\nmodels to automatically learn the syntactic structure of human\nlanguag
 es will impact upon both our understanding of how children\nlearn language
 \, and our ability to build sophisticated language\ntechnologies. In this 
 talk I will describe our recently developed\nstate-of-the-art approach to 
 syntax induction. Using hierarchical\nnon-parametric Bayesian priors we ha
 ve created probabilistic  models\nof syntactic part-of-speech and dependen
 cy grammar that are able to\nintegrate information across a range of granu
 larities. The promising\nresults achieved by these models indicate that th
 e great challenge of\nGrammar Induction may not be as intractable as long 
 thought.\n
LOCATION:Lecture Theatre 1\, Computer Laboratory
END:VEVENT
END:VCALENDAR