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/2008/newsitem/2541/22- October-2008-Computational-Linguistics-Seminar-Fed erico-Sangati DTSTAMP:20080929T000000 SUMMARY:Computational Linguistics Seminar, Federic o Sangati ATTENDEE;ROLE=Speaker:Federico Sangati DTSTART;TZID=Europe/Amsterdam:20081022T153000 DTEND;TZID=Europe/Amsterdam:20081022T000000 LOCATION:P.327, Euclides building, Plantage Muider gracht 24, Amsterdam DESCRIPTION:I will present several algorithms for assigning heads in phrase structure trees, based o n different linguistic intuitions on the role of h eads in natural language syntax. The starting poin t of the approach is the observation that a head-a nnotated treebank defines a unique lexicalized tre e substitution grammar. This allows us to go back and forth between the two representations, and def ine objective functions for the unsupervised learn ing of head assignments in terms of features of th e implicit lexicalized tree grammars. We evaluate algorithms based on the match with gold standard h ead-annotations, and the comparative parsing accur acy of the lexicalized grammars they give rise to. For more information and abstracts, see http:/ /www.illc.uva.nl/LaCo/CLS/ X-ALT-DESC;FMTTYPE=text/html:\n
\n I will present several algorithms for assigning he ads in phrase structure trees, based on different linguistic intuitions on the role of heads in natu ral language syntax. The starting point of the app roach is the observation that a head-annotated tre ebank defines a unique lexicalized tree substituti on grammar. This allows us to go back and forth be tween the two representations, and define objectiv e functions for the unsupervised learning of head assignments in terms of features of the implicit l exicalized tree grammars. We evaluate algorithms b ased on the match with gold standard head-annotati ons, and the comparative parsing accuracy of the l exicalized grammars they give rise to.\n
\n \n\n For more information an d abstracts,\n see http://www.illc .uva.nl/LaCo/CLS/\n
URL:/NewsandEvents/Archives/2008/newsitem/2541/22- October-2008-Computational-Linguistics-Seminar-Fed erico-Sangati END:VEVENT END:VCALENDAR