Sharing Metainformation to Guide Cooperative Search Among Heterogeneous Reusable Agents

A reusable agent is a self contained computational system that implements some speci c expertise and that can be embedded into diverse applications requiring that expertise Systems composed of heterogeneous reusable agents are potentially highly adaptable maintainable and a ordable assuming that integration issues such as in formation sharing coordination and con ict management can be e ectively addressed In this article we investigate the e ectiveness of sharing meta level search information to improve system performance speci cally with respect to how sharing a ects the quality of solutions and the runtime e ciency of a reusable agent system We rst give a formal description of shareable meta information in systems where agents have private knowledge and databases and where agents are speci cally in tended to be reusable We then present experimental results from a mechanical design system for steam condensers that demonstrate performance improvements related to information sharing and assimilation Finally we discuss the practical bene ts and limitations of information sharing in application systems comprising heterogeneous reusable agents Issues of pragmatic interest include determining what types of infor mation can realistically be shared and determining when the costs of sharing outweigh