论文部分内容阅读
知识库中存在的不一致性是影响整个系统性能的重要因素之一。在系统建立初期,由于知识库比较小,内容也比较简单,只要设计人员或知识工程师对初始知识模型进行反复推敲和精心构造,这类问题还是比较容易防止的。但是随着时间的推移,人们对知识库在最初设计阶段的种种考虑逐渐淡漠,同时由于知识库中新的知识不断加入,知识库越来越大,内容也越来越丰富复杂,这时知识库中各知识单元之间的相互影响和相互联系就随之变得复杂,难以跟踪和捉摸了。在这种情况下,知识库中的不一致性检测、识别和纠正显得更为重要。1、知识库的一致性问题知识库系统的组织按其结构可以划分为事实性知识(领域知识)和控制性知识,它
Inconsistencies in the knowledge base are one of the important factors that affect the overall system performance. In the initial stage of system establishment, because the knowledge base is relatively small and the contents are relatively simple, such problems can be easily prevented if designers or knowledge engineers repeatedly scrutinize and carefully construct the initial knowledge model. However, with the passage of time, people gradually indifferent to the consideration of the knowledge base in the initial design stage. At the same time, due to the continuous addition of new knowledge in the knowledge base, the knowledge base is getting larger and the content is more and more complex. At this time, knowledge The mutual influence and interrelationships between knowledge units in the library become complicated and hard to follow and elusive. In this case, inconsistency in the knowledge base to detect, identify and correct even more important. 1, the problem of consistency of knowledge base Knowledge base system according to its structure can be divided into factual knowledge (domain knowledge) and control knowledge, it