| 摘要: |
| 在KDD(knowledge discovery in database)中,对所发现的知识进行评价是一个很重要的环节.提出了一种针对KDD中因果关联规则的自动评价方法.该评价方法采用了全新的、有效的知识表示方法(语言场和语言值结构)和推理机制(因果关系定性推理机制),并且具有通用性和交互性的特征.给出了此评价方法的理论依据和构造过程,并提供了相应的算法.通过对具体实例的运行检验,证明了此评价方法的有效性.通过与相关工作的比较,证明了其先进性. |
| 关键词: 因果关联规则 感兴趣度 评价 KDD(knowledge discovery in database) |
| DOI： |
| 分类号: |
| 基金项目:国家自然科学基金资助项目(69835001);国家教育部科技重点项目([2000]175) |
|
| KDD中因果关联规则的评价方法 |
|
YANG Bing-ru,QI Yan-xia
|
| Abstract: |
| It is very important to evaluate the discovered rules in KDD (knowledge discovery in database). An evaluation method for causal rules is provided in this paper. The new and valid knowledge expression (language field and language value) and the reasoning mechanism (qualitative induction mechanism of causal relation) are used. The method is general and interactive. Its construction and the algorithm are given, and its validity is proved through case. By the comparison with the related work, it is proved to be an advanced method. |
| Key words: causal rules measures of interestingness evaluation KDD （knowledge discovery in database） |
