Intelligent Access to Data and Knowledge Bases via User's Topics of Interest

Retrieving relevant information in Data and Knowledge Bases containing a large number of di erent types of information is a non trivial problem. That is the reason why, in areas like Decision Support Systems, or Computer Aided Design, users need the help of Intelligent Systems. Such Intelligent Systems must be able to provide the users with useful information which is not explicitely requested. For this purpose, we consider a method where the user's needs are represented in terms of their topics of interest. This kind of user representation is integrated in an appropriate Modal logic, that allows the system to reason about the user's representation, the user's query and the DKB content, in order to determine the useful information. The main contribution of the paper is to present a rst attempt to the formalization of the concepts of topic, interesting topic and interesting piece of information. Several criteria are investigated for these de nitions, in particular to de ...

• Type: Conference
• Year: 1992
• Where: IFIP
• Authors: Sylvie Cazalens, Robert Demolombe
• Keywords: IFIP 1992, Intelligent Systems, Non Trivial Problem, Programming Languages, Useful Information