New methods and new systems are needed to filter or to selectively distribu
te the increasing volume of electronic information being produced nowadays.
An effective information filtering system is one that provides the exact i
nformation that fulfills user's interests with the minimum effort by the us
er to describe it. Such a system will have to be adaptive to the user chang
ing interest, In this paper we describe and evaluate a learning model for i
nformation filtering which is an adaptation of the generalized probabilisti
c model of Information Retrieval. The model is based on the concept of 'unc
ertainty sampling', a technique that allows for relevance feedback both on
relevant and nonrelevant documents. The proposed learning model is the core
of a prototype information filtering system called ProFile. (C) 1999 Elsev
ier Science Ltd. All rights reserved.