基于检索决策树的螨种分类鉴定专家系统

摘要

应用传统的分类检索方法进行粉螨亚目螨种的分类鉴定是一个复杂、繁琐的过程.随着人工智能技术的发展,采用专家系统技术编制螨种分类鉴定软件将改变传统的分类鉴定方式.通过学习粉螨亚目的分类学和专家系统的相关知识,将螨种的分类鉴定与专家系统有机地结合起来,采用面向对象方法,设计并实现了基于决策树的螨种分类鉴定专家系统.通过构建基于面向对象的决策树知识表示模型,提炼出螨种知识库的记录结构,根据节点对象的逻辑关系,设计了基于决策树的推理机制,并按照一定的推理控制策略实现了螨种的分类鉴定.结果证明,满足了用户的需要,具有重要的实践价值.%It is a complicated and tedious process to classify the mites by means of traditional methods. With the development of artificial intelligence,to design an expert system of classification and identification for Acaroid mites will definitely alter the methods of traditional classification. Through studying the relevant knowledge of taxonomy for the species of Acaroid Mites of Acaridida and expert systems, the characterization for mites and expert systems are combined. This paper applies an object-oriented method,designs and implements an expert system of classification and identification for Acaroid mites based on decision trees. According to the knowledge model of object-oriented decision trees, we propose the record strcture of the knowledge base for Acaroid mites, and design a reasoning mechanism based on decision trees by means of the logic of node objects. The reasoning mechanism realizs the identification of the mites by an inferential strategy. The experimental results prove that the system is of practical value, and satisfies the demand of users.