用数据挖掘方法扩充知识库的研究及应用

Research with Application of Expanding Knowledge Repository by Data Mining

讨论了数据挖掘与机器学习对于扩充知识库的异同 ,分析了知识库、数据库与数据挖掘在知识系统中的关系 .提出了一种基于 XML 的知识表示方法 XKR(XML- based Knowledge Representation) .XKR用 XML 作为统一的形式描述语言 ,把产生式、框架、语义网络、过程表示法等等多种传统的表示方法融合到一起 .由于 XML本身包含语义并能够无限扩充 ,所以 XKR可以描述不同背景不同类型的知识 ,实现知识融合 .通过应用实践发现 XKR知识库有优点也有缺陷 。

Data mining(DM) and machine learning(ML) are both popular term in knowledge system. They are very similar in some sense but not identical the same. We use data mining to expand a knowledge repository in order to exploit huge amount data in knowledge system. XKR(XML based Knowledge Representation), a unified formula description language based on XML, is presented to integrate several traditional knowledge representations such as production, frame, semantic net, process and so on. Because of XML can contain semantic itself, XKR is able to describe knowledge of various backgrounds and types so that we implemented knowledge fusion on it. With some applications we find highlights of XKR knowledge repository though there are drawbacks needed to improve.