综述:知识系统的V&V技术  

A Survey on Verification and Validation of Knowledge-based Systems

作　　者：曾庆田[1] 梁永全[1] 段华[1] 

机构地区：[1]山东科技大学信息科学与工程学院,青岛266510

出　　处：《计算机科学》2006年第2期19-24,共6页Computer Science

基　　金：自然科学基金的资助(#60073017和#60273019);科技部重大基础项目基金(#2001CCA03000和#2002DEA30036)的资助。

摘　　要：知识库的异常是影响整个知识系统性能的重要因素之一,因此必须对获取的知识进行校验。本文综述了知识库异常检测和验证的相关研究,给出了异常知识的分类及其危害性,分析了知识库验证困难的原因,介绍了用于知识库验证的静态和动态方法,列举了国际上几个著名的知识库验证工具,并对知识库验证的研究进行了展望。Knowledge abnormities including inconsistency, redundancy and incompletities are the main reasons to affect the knowledge-based system efficiencies,so the knowledge verification and validation （V＆V）are necessary. The related works about knowledge V＆V are summarized in this paper. The knowledge abnormities are classified and their harms to knowledge-based system are discussed. However, the knowledge V＆V are usually difficult since knowledge repre- sentation, update and scale. The static and dynamic methods for knowledge V＆V are listed respectively, and some known tools for knowledge V＆V are compared. Finally,tile future research for knowledge V＆V is discussed.

关 键 词：知识系统 知识异常 知识表示 知识验证 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] TP393.08[自动化与计算机技术—控制科学与工程]