With the development of network technology, the spread of information on Internet becomes more andmore quick. There are many types of complicated data in the information ocean. How to acquire useful knowledgequickly from the information ocean is the very difficult. The Text Mining based on Web is the new research fieldwhich can solve the problem effectively. In this paper, we present a structure model of Text Mining and research thecore arithmetic - Classification arithmetic. We have developed the Text Mining system based on Web and appliedit in the modern long-distance education. This system can automatically classify the text information of education fieldwhich is collected from education site on Internet and help people to browser the important information quickly andacquire knowledge.
本文在基于数据库与知识库协同机制的综合型知识发现系统KD(D&K)的基础上,首次提出基于知识发现系统的广义诊断型专家系统GDESKD(Generalized Diagnosis Type Expert System Based on Knowledge Discovery)构造方法。本文提出了GDESKD的总体结构框架,描述了其中的一些子系统;着重讨论了基于双库协同机制的动态知识库和其中的各知识库的形成过程;初步演示试验的结果表明GDESKD构造是有效与可行的。GDESKD是一种具有全新特征的专家系统,可能形成新的研究方向,它为解决知识库中知识的丰富性提供了极大的可能。
