Agent for information source location in a dynamic DSS

Abstract

Inside an organization there is information that can only be generated by people, then, when a decision maker in any time and anywhere of an organization requires this information, he/she has to solicit it to who has the ability of generate it. To fulfill this information requirement, we are developing a multiagent system that we call dynamic Decision Support System (DSS). This system is composed by Domain Representative Agents (DRAs); an Information Source Locator Agent (ISLA) and mobile agents called Query Coordinator Agents (QCAs). The ISLA is the agent responsible for guiding information requirements from users to distributed DSS domains that offer greater possibilities of answering them. It has the ability to interpret queries formulated in natural language, to identify the relationships among the characteristics of queries and domains, and to learn from errors that it make during its operation to diminish the number of consulted domains in each information requirement. The purpose of this work is to describe the ISLA architecture and its main components; using design patterns and knowledge engineering methodologies.