Toward a Methodology for Utilizing Sources of Evidence in Relevance Feedback
Emanuele Di Buccio
University of Padova
Wednesday 3 November 2010 at 11:30 AM
Salle 3195, Pavillon André-Aisenstadt
The variety of user intents and interests is a key factor which makes information retrieval challenging, since textual queries, as a mostly adopted representation of information need, are often insufficient to describe what a user perceives as useful for his information goal. Other evidence can be adopted to complement this representation. But the utilization of multiple and diverse sources of evidence within a document ranking function increases the complexity of its design because of the diversity among sources. This talk illustrates a methodology which seamlessly describes the sources and the document ranking function in order to reduce this complexity. In particular the methodology models the contribution of diverse sources as vector subspaces and each document as a vector. The distance between the document vector and the subspace is adopted as measure of the degree to which the document satisfies the dimension of the information need representation corresponding to the considered source. Moreover, this talk discusses an utilization of the methodology to obtain and exploit a user behavior model built from the observed user interaction features for representing the user’s information need and accessed documents.
To receive weekly talk announcements, please send an e-mail to firstname.lastname@example.org. Simply write a message containing the single line 'subscribe ralli' (without the quotes).
See all the weekly talks for the year:1991 1992 1993 1994 1995 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021