BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Modeling and Evaluating Information Retrieval Results - Evangelos 
 Kanoulas\, University of Sheffield
DTSTART:20110411T094000Z
DTEND:20110411T104000Z
UID:TALK30649@talks.cam.ac.uk
CONTACT:Microsoft Research Cambridge Talks Admins
DESCRIPTION:In this talk I will discuss two different pieces of work I hav
 e done in the field of Information Retrieval. In the first part of my talk
  I will focus on my work on modeling score distributions produced by infor
 mation retrieval systems. Information retrieval systems assign scores to d
 ocuments according to some definition of relevance to a user's request and
  return documents in a descending order of their scores. Given this ranked
  list of documents and their corresponding scores\, inferring the score di
 stributions of relevant and non-relevant documents is an essential task fo
 r numerous information retrieval applications\, such as information filter
 ing\, topic detection\, meta-search\, and distributed IR. Modeling score d
 istributions in an accurate manner is often the basis of such inferences. 
 In this part of my talk I will revisit the choice of distributions used to
  model documents' scores. First\, I will discuss some assumptions and intu
 itions behind modeling score distributions. Then I will present a better m
 odel for score distributions directly dictated by the data\, using a riche
 r class of density functions than the ones dominating the literature and a
 pplying Variational Bayes to automatically trade-off the goodness-of-fit a
 nd the model complexity.
LOCATION:Small lecture theatre\, Microsoft Research Ltd\, 7 J J Thomson Av
 enue (Off Madingley Road)\, Cambridge
END:VEVENT
END:VCALENDAR