BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Adjective modification in compositional distributional semantics -
  Eva Maria Vecchi\, Computer Lab\, Cambridge
DTSTART:20140411T110000Z
DTEND:20140411T120000Z
UID:TALK50606@talks.cam.ac.uk
CONTACT:Tamara Polajnar
DESCRIPTION: In this talk\, I discuss the ability of compositional distrib
 utional semantics to model adjective modification. I present three studies
  that explore the degree to which semantic intuitions are grounded in the 
 distributional representations of adjective-noun phrases\, as well as prov
 ide insight into various linguistic phenomena by extracting unsupervised c
 ues from these distributional representations. First\, I investigate degre
 es of adjective modification: intersectively used color terms\, subsective
 ly used color terms\, and intensional adjectives. Next\, I propose an appr
 oach to characterize semantic deviance of composite expressions using dist
 ributional semantic methods. I present a set of simple measures extracted 
 from distributional representations of words and phrases\, and show that t
 hey are more significant in determining the acceptability of novel adjecti
 ve-noun phrases than measures classically employed in studies of compound 
 processing. Finally\, I use compositional distributional semantic methods 
 to investigate restrictions in adjective ordering. Specifically\, I focus 
 on properties distinguishing adjective-adjective-noun phrases in which the
 re is flexibility in the adjective ordering from those bound to a rigid or
 der. I explore a number of measures extracted from the distributional repr
 esentation of such phrases which may indicate a word order restriction. Ov
 erall\, this work provides strong support for compositional distributional
  semantics\, as it is able to generalize and capture the complex semantic 
 intuition of natural language speakers for adjective-noun phrases\, even w
 ithout being able to rely on co-occurrence relations between the constitue
 nts.
LOCATION:FW26\, Computer Laboratory
END:VEVENT
END:VCALENDAR
