Modeling a Multi-Agent System for Retrieving Information from Distributed Sources

Abstract

This paper presents a multi-agent system called ISAME and designed for intelligent information retrieval from heterogeneous distributed sources. ISAME constitutes a virtual library that supplies a set of software agents with a simplified access to a set of dynamic information sources available under electronic formats, as well as services for facilitating and optimizing information retrieval. The system also uses TCP/IP communication protocols, and proposes a series of garbage-collection mechanisms to avoid the preservation and propagation of information among agents, or messages that become inaccessible or outdated as well as the use of the resources that become undesirable. The originality of this system rests on the fact that it simplifies the information retrieval from distributed heterogeneous sources by making these sources transparent to the user.