| The expert system (ES) is a kind of software system which includes lots of knowledge and experience in especial fields. It applies Artificial Intelligence (AI) technology and simulates the thinking process of human experts to solve problems in some fields, the capability of which can approach and even exceed that of human experts. ES becomes one of the most active fields of AI research, which accelerates development of the basic theory and technology of AI. The kindred relationship (KR) contacts closely with people?ˉs social life, and there are all kinds of expressions of KR. This paper studies the reasoning of KR based on ES, and emphasizes how to correctly acquire knowledge from a sentence which comes from a user, analyze the intention of the user, and answer all kinds of questions from the user. The author has some creative works as follows: According to common KR, the author proposes five KR gene predicates which decide KR and still can express both relationship of individuals and KR knowledge. Also about sixty predicate rules are brought forward. A method of KR knowledge acquisition is introduced according to expression forms of KR, which includes both automatic knowledge acquisition and non-automatic knowledge acquisition. The acquired KR knowledge is changed into KR gene predicates instances and conclusion predicates instances. Self-learning strategy is provided. The writer designs KR reasoning mechanism, advances three kinds of efficient arithmetic according to common reasoning types: positive link reasoning, negative link reasoning, combined link reasoning. These arithmetic can solve many problems. The KR knowledge can be stored into knowledge base, and the paper designs the structures of some KR knowledge bases and puts up the model of KR logical reasoning expert system and computer-processing.
|
PDF Full Text Request