论文部分内容阅读
本文对专家系统的学习问题进行了一些探讨,并用LISP语言实现了一个具有一定学习能力的专家系统。该系统除了具有推理判断功能外,还具有教授示例学习功能和自动扩展知识库、整理知识库等功能。本文以动物识别规则库为例介绍了专家系统的学习功能的实现方法。若配以不同的规则库,则本系统的程序可以适用于不同的产生式专家系统。本系统一经建立,以后就根据使用中人机对话的不同,各个专家系统自行发展和扩充,提高智能水平。本系统是用LISP语言编写,在PDP11/23+机上实现的。
In this paper, some problems about the learning of expert system are discussed, and an expert system with some learning ability is realized by LISP language. The system not only has the function of reasoning and judgment, but also has the function of teaching example learning and automatically expanding the knowledge base and arranging the knowledge base. In this paper, we take the animal identification rule base as an example to introduce the realization method of the learning function of expert system. If accompanied by different rules of the library, the system of programs can be applied to different production expert system. Once the system is established, the various expert systems will develop and expand on their own according to different man-machine conversations in the future and improve the intelligence level. The system is written in LISP language, implemented on the PDP11 / 23 + machine.