BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Drupal iCal API//EN
X-WR-CALNAME:Events items teaser
X-WR-TIMEZONE:America/Toronto
BEGIN:VTIMEZONE
TZID:America/Toronto
X-LIC-LOCATION:America/Toronto
BEGIN:DAYLIGHT
TZNAME:EDT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
DTSTART:20130310T070000
END:DAYLIGHT
BEGIN:STANDARD
TZNAME:EST
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
DTSTART:20131103T060000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
UID:69b473ed1f384
DTSTART;TZID=America/Toronto:20140226T110000
SEQUENCE:0
TRANSP:TRANSPARENT
DTEND;TZID=America/Toronto:20140226T110000
URL:https://uwaterloo.ca/centre-pattern-analysis-machine-intelligence/event
 s/invited-talk-modeling-term-associations-probabilistic
LOCATION:Centre for Environmental and Information Technology 200 University
  Ave W EIT-3142 Waterloo ON N2L3G1 Canada
SUMMARY:Invited Talk: Modeling Term Associations for Probabilistic Informat
 ion\nRetrieval
CLASS:PUBLIC
DESCRIPTION:MODELING TERM ASSOCIATIONS FOR PROBABILISTIC INFORMATION RETRIE
 VAL\n\nABSTRACT:\n\nTraditionally\, in many probabilistic retrieval models
 \, query terms are\nassumed to be independent. Although such models can ac
 hieve reasonably\ngood performance\, associations can exist among terms fr
 om human\nbeing.s point of view. There are some recent studies that invest
 igate\nhow to model term associations/dependencies by proximity measures.\
 nHowever\, the modeling of term associations theoretically under the\nprob
 abilistic retrieval framework is still largely unexplored. In this\ntalk\,
  I will introduce a new concept named Cross Term\, to model term\nproximit
 y\, with the aim of boosting retrieval performance. With Cross\nTerms\, th
 e association of multiple query terms can be modeled in the\nsame way as a
  simple unigram term. In particular\, an occurrence of a\nquery term is as
 sumed to have an impact on its neighboring text. The\ndegree of the query 
 term impact gradually weakens with increasing\ndistance from the place of 
 occurrence. We use shape functions to\ncharacterize such impacts. Based on
  this assumption\, we first propose\na bigram CRoss TErm Retrieval (CRTER2
 ) model as the basis model\, and\nthen recursively propose a generalized n
 -gram CRoss TErm Retrieval\n(CRTERn) model for n query terms where n > 2. 
 \nSpecifically\, a bigram Cross Term occurs when the corresponding query\n
 terms appear close to each other\, and its impact can be modeled by the\ni
 ntersection of the respective shape functions of the query terms. For\nn-g
 ram Cross Term\, we develop several distance metrics with different\nprope
 rties and employ them in the proposed models for ranking. We also\nshow ho
 w to extend the language model using the newly proposed cross\nterms. Exte
 nsive experiments on a number of TREC collections\ndemonstrate the effecti
 veness of our proposed models. \n\nBIOGRAPHY:\n\nJIMMY HUANG is a Professo
 r & Director at the School of Information\nTechnology and the founding dir
 ector of Information Retrieval &\nKnowledge Management Research Lab at the
  York University.
DTSTAMP:20260313T203037Z
END:VEVENT
END:VCALENDAR