基于领域本体与概念格的关联规则挖掘

关联规则是数据挖掘中的核心任务之一,近年来国内外对关联规则算法的改进取得了比较大的成果.概念格是由二元关系导出的形式化工具.体现了概念内涵和外延的统一,非常适合于发现数据中的潜在关系,因此关联规则的提取也是概念格的一个主要的应用领域,极大的提高了关联规则的挖掘效率,然而由于缺乏领域知识的指导,所挖掘出的规则有些是无意义的或无法满足用户的需要,所以在规则的提取中需要引入领域知识,而领域本体是领域知识的清晰而结构化的表示,因此提出了应用领域本体对生成的概念格进行调整,从而实现对规则提取的指导,以发掘出高层关联规则以及多层次间的关联规则,以满足用户的需要.

Mining association rules on concept lattice using domain ontology

Association rules mining is one of the main tasks in data mining.Recently,association rules algorithm has been improved greatly.Concept lattice is used as a powerful formal concept analysis tool to reveal the accordance of extensions and intensions.It is very fit for discovering potential relations in database.Efficiency has greatly improved by using concept lattice in mining association rules in data mining.But because of the absence of domain knowledge,many useless rules which cannot satisfy the users' requirement are generated.A thought that using domain ontology in formal concept analysis to direct association rules mining is proposed here.There-fore,high level association rules and multiple level association rules can be mined here.