Distributed Knowledge Base Systems for Diagnosis and Information Retrieval

摘要

This report summarizes work done in artificial intelligence. Research is reported on in the following areas: applications of generic task methodology to learning, design, and planning; theoretical and critical studies of connectionism, neural nets, and symbolic computation; and studies of the uses of distributed systems for concurrent algorithms/implementation for abductive assembly a basic generic problem solving task. Topics include: A corrective learning procedure using different explanatory types; A task oriented approach to knowledge-based systems for process engineering design; An information processing model of japanese foreign and energy policy decision making; Understanding device feedback; Connectionism and information processing abstractions--The message still counts more than the medium; Some experiments with neural nets; From numbers to symbols to knowledge structures--AI perspectives on the classification task; Distributed synthesis of composite explanatory hypotheses.