Deterministic Strategies for Query (re)formulation in Information Retrieval

Relevance feedback is a process whereby a user examines the documents selected by a retrieval system and provides feedback t o the system as t o their relevance. Such a feedback can then be used t o formulate an optimal query with respect t o the current infwmation need of the user. This process of query (re)formulation can be based on probabilistic concepts, where Bayesian decision theory provides the framework for a decision rule, or ideas which instead employ deterministic strategies. The former class of techniques are limited by the fac t tha t (strong) assumptions have t o be made concerning the nature of the conditional orobabilitv densitv functions characterizinn the data. In contrast, deterministic techniques, which do not require any explicit assumptions about the distribution of the various descriptor vaiues, can b e adopted. Such methods would have t h e advantages of being %on-parametric" and robust (useful in a wide variety of contexts). A d a s s of deterministic techniques has been advocated by Salton and the SMART project group at Cornell. However, tha t approach d w s not obtain a new q w r y t h a t can be claimed t o be optlmal in a certain sense. In this work, a deterministic method t h a t obtains an optimal query according t o a prescribed criterion is advanced. Furthermore, i t will be demonstrated t h a t such methods a r e applicable not only when there a r e two classes of relevance (relevant and non-relevant) but also when t h e feedback distinguishes documents according t o several degrees of relevance.