Knowledge Structural Modeling for Developing Expert Systems.

知识获取是专家系统开发的瓶颈。本研究的目的是应用柔性解释结构建模（FISM）这一构造系统结构模型的方法，开发一种新的有效的知识获取方法。这项研究计划从1990年至1991年为期两年。第一阶段的主要研究结果如下：1.对知识获取方法进行了综述。重点讨论了个人建构理论在知识获取系统ETS中的应用.第二阶段的研究主要包括以下几个方面：1.将FISM的功能扩展到处理多维关系和关系度。引入模糊二元关系来考虑关联度; 2.通过引入互连关系矩阵处理多维关系;在AS 4330工作站上实现了测试系统。研究成果已应用于计算机语言的自动推理、定量推理和句法学习等领域，但如何处理基于实例的推理仍是一个有待解决的问题。

Knowledge acquisition is the bottle neck of developing experts systems. This research project aimed to develop a new effective method for knowledge acquisition by applying Flexible Interpretive Structural Modeling (FISM) Which is a method used to construct systems structural modeling. The research project was planed for a period of two years, from 1990 to 1991. The main results of the first period are as follows :1. Many knowledge acquisition methods were surveyed. Especially, Personal construct theory which is used in a knowledge acquisition system ETS was discussed,2. Extension of functions of FISM to treat multidimensional relation and degrees of relation were considered.The results of the second period are summarized as fololws :1. Fuzzy binary relation was introduced for considering the degree of relation,2. Multidimensional relation was treated by introducing the interconnection relation matrix,3. A test system was implemented on an AS4330 work station,4. Results were applied to automated reasoning, quantitative reasoning and syntax learning of computer languages.The problem of how to deal with the case based reasoning hereafter still remains.