| This research work attempts to address some issues of the Web information retrieval problem. The explosive growth of information on the Web combined with the diverse users interests have caused a failure of existing tools to search the Web and motivated for effective alternatives. The strategy pursued to address these challenges is to have a distributed group of intelligent agents that are goal and pattern driven, autonomous, and rational in a cooperative and interactive setting. The system is capable of satisfying user needs, and shall introduce novel methodologies in receiving queries, retrieving answers, and presenting results. A multi-tier agent-based architecture in which a mediator agent fuses decisions collected from decision makers is presented. An intelligent fusion technique, the team consensus approach, is used to fuse agent decisions on the relevancy of retrieved information. The methodology models the interaction and brings team into consent. This thesis presents the proposal, and provides comparative analysis on results collected from a realized prototype of the system. |
