BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:A Hierarchical Bayesian Language Model based on Pitman-Yor Process
 es - Matt Shannon (University of Cambridge)
DTSTART:20090122T140000Z
DTEND:20090122T153000Z
UID:TALK15404@talks.cam.ac.uk
CONTACT:Shakir Mohamed
DESCRIPTION:I will be discussing:\n\n* A Hierarchical Bayesian Language Mo
 del based on Pitman-Yor Processes\, Yee Whye Teh\, http://www.gatsby.ucl.a
 c.uk/~ywteh/research/bayesnlp/acl2006.pdf\n\nN-gram language modelling tra
 ditionally uses some form of "smoothing" technique to allocate some probab
 ility mass to unseen N-grams.  Over the years people have come up with smo
 othing schemes that perform pretty well\, but it's not easy to get a handl
 e on what they're doing\, and how to improve them.\n\nIn this paper\, Teh 
 shows that a hierarchical Bayesian language model with a very simplistic m
 odel of context performs pretty much as well as the current state of the a
 rt smoothing schemes\, and in fact has strong similarities to an existing 
 smoothing scheme.\n
LOCATION:Engineering Department\, CBL Room 438
END:VEVENT
END:VCALENDAR