Research And Realization On The Knowledge Representation Of Soldier Psychology Consultation Expert System

To meet the situation of the military informationization, and to satisfy the realistic need of the military psychological work in the new period, this paper research the issues relevant to the problem that how to construct a Soldiers Psychological Counseling Expert System (SPCES) with mature technology of artificial intelligence and expert system, and highlight the soldiers psychological counseling modeling, knowledge representation, knowledge base realization and maintenance, and other aspects of this system.1. We discuss the development status of artificial intelligence and expert system, the realistic need of soldier psychological counseling, and propose the researching and developing background of SPCES. We also introduce SPCES briefly.2. Domain modeling. We analyze and sort out the knowledge in the field of soldier psychological counseling provided by experts, research the theory model, diagnostic model, intervention model of this domain. We also establish a soldier psychological counseling domain model, identify the basic concepts, objects and relations, and summarize the strategic knowledge and the uncertainties of knowledge in the area. Our work fills the domestic gap of modeling research in the field of psychological counseling and also laid the foundation for researching and developing SPCES.3. Knowledge representation. Concerning the characteristics of various kinds of knowledge in soldier psychological counseling area, we extend the weighted fuzzy production rules, and propose a new extended weighted fuzzy production rules to make it more effective in representing the SPCES knowledge. On the basis of study knowledge representation that represents different types of knowledge in this area, we construct the knowledge base for SPCES.4. Maintenance of the knowledge base. We propose a general detection and refinement algorithm for SPCES knowledge base, and design a method of knowledge maintenance which is based on the explain mechanism.5. Knowledge acquisition. We analyze the difficulty of knowledge acquisition, and design and realize a non-automatic knowledge acquisition method , knowledge editor. We conduct a preliminary study of CBR, which is an automatic knowledge acquisition method that is base on machine learning. At last, we give the SPCES knowledge acquisition subsystem program.