基于CBR的羊病诊断专家系统分析与研究  

Research on Sheep&Goat Diagnosis Expert System Based on CBR

作　　者：聂艳召[1] 李书琴[2] 

机构地区：[1]西北农林科技大学信息工程学院,陕西杨凌712100 [2]安阳师范学院,河南安阳455000

出　　处：《农机化研究》2007年第7期33-36,共4页Journal of Agricultural Mechanization Research

基　　金：国家"863"资助项目(2002AA2Z4051);西北农林科技大学人才专项基金项目(04ZR015)

摘　　要：鉴于基于规则的专家系统在知识获取上存在的瓶颈问题,分析研究了基于案例推理的羊病诊断专家系统关键问题的设计思想和实现方法。同时,针对羊病症状信息的特点,提出了建立症状索引表描述字符型数据、特征权值反映症状对案例的影响;采用特征向量空间的加权检索算法,提出了症状特征相似性计算方法。采用SQLServer数据库管理和Java语言进行系统实现,并通过实例进行了验证,达到了设计要求,对生产实践具有指导价值。In view of the fact that the rule-based expert system encounters the bottleneck problem which exists in gaining the knowledge, this paper researches on the critical problem in case-based reasoning expert system for sheep ＆ goat diseases diagnosis. According to the feature of sheep ＆ goat diseases, it brings up a method of building a symptom index table to describe char-type data, and using character-weight to reflect the importance of each symptom in one case, using nearest neighbor retrieving strategy to calculate the similarity between two cases. The system is fulfilled with SQL Server plus Java, and it is proved meets the designing request, and it is has the significance to the production practice.

关 键 词：畜牧学 专家系统 理论研究 案例推理 案例表示 案例检索 

分 类 号：S818.9[农业科学—畜牧学] TP182[农业科学—畜牧兽医]