BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:reading-group: Interpolating Between Types and Tokens by Estimatin
 g Power-Law Generators - Speaker to be confirmed
DTSTART:20060119T110000Z
DTEND:20060119T120000Z
UID:TALK4681@talks.cam.ac.uk
CONTACT:David MacKay
DESCRIPTION:http://cog.brown.edu/~gruffydd/papers/typetoken.pdf\n\nPaper-a
 bstract:\nStandard statistical models of language fail to capture one of t
 he most\nstriking properties of natural languages: the power-law distribut
 ion in\nthe frequencies of word tokens. We present a framework for develop
 ing\nstatistical models that generically produce power-laws\, augmenting s
 tandard\ngenerative models with an adaptor that produces the appropriate\n
 pattern of token frequencies. We show that taking a particular stochastic\
 nprocess  the Pitman-Yor process  as an adaptor justifies the appearance\n
 of type frequencies in formal analyses of natural language\, and improves\
 nthe performance of a model for unsupervised learning of morphology.
LOCATION:Room 911\, Rutherford Building\, Cavendish Laboratory\, Departmen
 t of Physics
END:VEVENT
END:VCALENDAR