Modeling intelligent information resource agents for dynamic heterogeneous sources

To discover and retrieve information in an open and dynamically changing environment, a multiagent system, called ISAME, has been proposed. The information gathering and analysis, in this system, are performed over networks of autonomous heterogeneous information sources. ISAME serves as a facilitating environment for four categories of agents, each one being specialized in a series of activities needed to complete a main goal which is performing the role of a digital version of user assistant dedicated for intelligent information retrieval. Among those agent categories, Information Agents (IA) present some important conceptual challenges. This paper presents a generic object-oriented model for IA that respects the dynamical and heterogeneous nature of data sources, and fits the ISAME architecture specifications.