BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Subjectivity Recognition on Word Senses - Katja Markert - Universi
 ty of Leeds
DTSTART:20100219T120000Z
DTEND:20100219T130000Z
UID:TALK22722@talks.cam.ac.uk
CONTACT:Laura Rimell
DESCRIPTION:In the field of opinion mining\, considerable work has gone in
 to the creation of word lists annotated for subjectivity\, i.e. marking wo
 rds such as "positive\, beautiful\, crank" as subjective and others such a
 s "chair\, woman\, read" as (mostly) objective. These word lists are then 
 used in the automatic identification of opinions at the sentence or docume
 nt level.\n\nHowever\, a variety of words are subjectivity-ambiguous\, i.e
 . they have at least one objective and one subjective sense as is shown by
  the two example senses of "positive" below.\n\n(1) OBJECTIVE: positive\, 
 electropositive---having a positive electric charge\; protons are positive
  (2) SUBJECTIVE plus\, positive---involving advantage or good\; a plus (or
  positive) factor\n\nIn this talk\, I concentrate on this latter problem b
 y automatically creating lists of subjective senses\, instead of subjectiv
 e words\, via adding subjectivity labels for senses to electronic lexica\,
  using the example of WordNet. Specifically\, I will discuss the following
  results:\n\n(1) Human annotation experiments which show that assigning su
 bjectivity to word senses is a well-defined task. These experiments also s
 how that subjectivity-ambiguity is frequent.\n\n(2) A semi-supervised appr
 oach based on minimum cuts that assigns subjectivity labels to word senses
  in lexical relation graphs.\n\nThis algorithm outperforms supervised grap
 h and non-graph based algorithms significantly\, reducing the error rate b
 y up to 40%. In addition\, the semi-supervised approach achieves the same 
 results as the supervised framework with less than 20% of the training dat
 a.\n\n[This talk has been postponed from 22 January.]
LOCATION:SW01\, Computer Laboratory
END:VEVENT
END:VCALENDAR