CAER: AN ONTOLOGY-BASED COMMUNITY OF AGENTS FOR EMERGENCY RELIEF

摘要

With the recent explosion in Web usage, increased attention has been placed on the difficult challenges of managing and manipulating widely diverse sources of information. Rather than continue on a course of developing special purpose decision support systems and then devising techniques for pulling Web data into them, we have proposed the opposite approach, designing decision support systems as small additions to a larger Web environment Focusing on an emergency command center example, we have proposed an integrated system architecture that mirrors the Web. Information agents are then proposed to provide a layer that tailors retrievals to reflect user and application driven preferences. A supervisor agent takes on the task of identifying relevant data sources and, when necessary, translating the query to terms that are compatible with those used by an outside data source. The translation is accomplished through the use of an application-oriented ontology. This approach facilitates access to the Internet while supporting internally developed data structures associated with decision support tools.