中文摘要：

针对关联规则之间存在的冗余性问题,已提出多种精简关联规则模型,但这些模型仍不同程度存在紧致度欠佳、信息丢失或恢复算法复杂的问题。提出了一种含更丰富关联信息的基本关联规则,并以基本关联规则为基础构建无损的精简关联规则集合,它是原始关联规则集合的子集,并能据此完全恢复原始关联规则集合。给出了基本关联规则模型的定义,证明了该精简模型的几个重要性质,并设计了用于挖掘该类规则的挖掘算法。实验表明,基本关联规则模型比现有的关联规则精简模型更加紧致。

英文摘要：

In view of the redundancy among the association rules mined from dataset, a variety of concise representatives have been suggested for representing the whole raw association rule set. But there more or less exist deficiencies in these models, such as poor compactness, information loss or complex recovery algorithm and etc. This paper proposed a new concise associa- tion rules representation model based on basic association rule, named as BAR, which contains more rich relation information to construct a lossless representative of raw association rule set. This association rule representative, which was a subset of the raw association rules, could be easily used to restore the whole raw association rules. Definition of BAR was clearly defined and several proposition and theories related to BAR are proved too. Finally designed a BAR mining algorithm based on lexieo- graphic rule tree. The experiments show that the representative model based on BAR is lossless and is more compact than other existing concise association rules representatives.