RETRIEVAL AND REASONING IN DISTRIBUTED CASE BASES

The proliferation of electronically available networked information ha s led researchers to examine the issues involved in developing automat ed methods for gathering information in response to a query from a use r. However, most of this literature deals with locating, gathering, an d selecting the best response to a query from among a multitude of res ponses from different repositories or digital libraries. This paper de als with a different model of response to a query, involving compositi on of mutually related partial responses spread across a network of in formation repositories. We present a system for cooperative retrieval and composition of a case in which subcases are distributed across dif ferent agents in a multiagent system. From a Gestalt perspective, a go od overall case may not be the one derived from the summation of best subcases. Each agent's local view may result in best local cases, whic h when assembled may not result in the best overall case in terms of g lobal measures. We propose a negotiation-driven case retrieval algorit hm as an approach to dynamically resolving inconsistencies between dif ferent case pieces during the retrieval process. (C) 1996 Academic Pre ss, Inc.