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/2017/newsitem/8687/23- March-2017-Computational-Linguistics-Seminar-Raffa ella-Bernardi DTSTAMP:20170320T141501 SUMMARY:Computational Linguistics Seminar, Raffael la Bernardi ATTENDEE;ROLE=Speaker:Raffaella Bernardi (Trento) DTSTART;TZID=Europe/Amsterdam:20170323T170000 LOCATION:Room F3.20, Science Park 107, Amsterdam DESCRIPTION:Linguistics quantifiers have been the realm of Formal Semantics. A lot is known about th eir formal properties and how those properties aff ect logical entailment, the licensing of polarity item, or scope ambiguities. Less is known about ho w quantifiers are acquired by children and even le ss about how computational models can learn to qua ntify objects in images. In this talk, we will rep ort on our findings in this direction. First of al l, we will explain why the task is interesting and challenging for a Language and Vision model. Seco ndly, we will report our evaluation of state-of-th e-art neural network models against this task. Thi rdly, we will compare the acquisition of quantifie rs with the acquisition of cardinals. We will show that a model capitalizing on a `fuzzy' measure of similarity is effective for learning quantifiers, whereas the learning of exact cardinals is better accomplished when information about number is pro vided. X-ALT-DESC;FMTTYPE=text/html:\n
Linguistics qu antifiers have been the realm of Formal Semantics. A lot is known about their formal properties and how those properties affect logical entailment, th e licensing of polarity item, or scope ambiguities . Less is known about how quantifiers are acquired by children and even less about how computational models can learn to quantify objects in images. I n this talk, we will report on our findings in thi s direction. First of all, we will explain why the task is interesting and challenging for a Languag e and Vision model. Secondly, we will report our e valuation of state-of-the-art neural network model s against this task. Thirdly, we will compare the acquisition of quantifiers with the acquisition of cardinals. We will show that a model capitalizing on a `fuzzy' measure of similarity is effective f or learning quantifiers, whereas the learning of e xact cardinals is better accomplished when informa tion about number is provided.
URL:http://www.illc.uva.nl/LaCo/CLS/ END:VEVENT END:VCALENDAR