BEGIN:VCALENDAR VERSION:2.0 PRODID:ILLC Website X-WR-TIMEZONE:Europe/Amsterdam BEGIN:VTIMEZONE TZID:Europe/Amsterdam X-LIC-LOCATION:Europe/Amsterdam BEGIN:DAYLIGHT TZOFFSETFROM:+0100 TZOFFSETTO:+0200 TZNAME:CEST DTSTART:19700329T020000 RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU END:DAYLIGHT BEGIN:STANDARD TZOFFSETFROM:+0200 TZOFFSETTO:+0100 TZNAME:CET DTSTART:19701025T030000 RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU END:STANDARD END:VTIMEZONE BEGIN:VEVENT UID:/NewsandEvents/Archives/2011/newsitem/4057/1-S eptember-2011-Computational-Linguistics-Seminar-Su zanne-Stevenson DTSTAMP:20110831T000000 SUMMARY:Computational Linguistics Seminar, Suzanne Stevenson ATTENDEE;ROLE=Speaker:Suzanne Stevenson (Toronto) DTSTART;TZID=Europe/Amsterdam:20110901T150000 DTEND;TZID=Europe/Amsterdam:20110901T000000 LOCATION:Room D1.113, Science Park 904, Amsterdam DESCRIPTION:Early verb learning in children seems an almost miraculous feat. In learning a verb, chi ldren must learn both the basic meaning of the eve nt ("falling" or "eating"), as well as the allowab le structures in their language for correctly comm unicating the participants in that event ("The gla ss fell", but not "She fell the glass"). Moreover, given the sparsity of evidence, children must be able to abstract away from specific usages they ob serve in order to use their knowledge of verbs pro ductively. Finally, children must accomplish all t his in the face of a high degree of variability am ong verbs, along with much noise and uncertainty i n the input data, and with no explicit teaching. D o children require innate knowledge of language to accomplish this, or are general cognitive learnin g mechanisms sufficient to the task? We have devel oped various computational models of verb learning using unsupervised clustering over simple statist ical properties of verb usages. Our findings suppo rt the claim that general learning mechanisms are able to acquire abstract knowledge of verbs and to generalize that knowledge to novel verbs and situ ations. This is joint work with Paola Merlo, Afra Alishahi, and Chris Parisien. For more informati on and abstracts, see http://www.illc.uva.nl/LaCo/ CLS/. X-ALT-DESC;FMTTYPE=text/html:\n
Early ve rb learning in children seems an almost miraculous \n feat. In learning a verb, children must learn both the basic\n meaning of the event ("falling" or\n "eating&quo t;), as well as the allowable structures in\n their language for correctly communicating the participants in\n that event ("The gla ss fell", but not "She fell\n the glass"). Moreover, given the sparsity of evi dence,\n children must be able to abstract away from specific usages\n they observe in order to use their knowledge of verbs\n pr oductively. Finally, children must accomplish all this in\n the face of a high degree of vari ability among verbs, along\n with much nois e and uncertainty in the input data, and with no\n explicit teaching. Do children require inn ate knowledge of\n language to accomplish t his, or are general cognitive learning\n me chanisms sufficient to the task? We have developed various\n computational models of verb lea rning using unsupervised\n clustering over simple statistical properties of verb\n usa ges. Our findings support the claim that general l earning\n mechanisms are able to acquire ab stract knowledge of verbs and\n to generali ze that knowledge to novel verbs and situations. p>\n
This is joint work with Paola Merlo , Afra Alishahi, and\n Chris Parisien.
\ n \nFor more information and abstrac ts, see\n http://www.illc.uva.nl/L aCo/CLS/.
URL:/NewsandEvents/Archives/2011/newsitem/4057/1-S eptember-2011-Computational-Linguistics-Seminar-Su zanne-Stevenson END:VEVENT END:VCALENDAR