一种近似概念格及其在知识发现中的应用研究*

针对经典概念格基于完备形式背景,在缺值形式背景下不能很好的获取关联规则知识的问题,在偏小近似形式背景的基础上,提出了一种τ-近似概念格来解决信息不完备背景下的知识发现问题,利用它能够获取精度和可信度较高的关联规则知识。文章对τ-近似概念格的构建及其关联规则提取过程进行了详细描述,并将其与偏小近似概念格和偏大近似概念格这两种解决策略进行了案例分析对比,验证了它的可行性和有效性。

An Approximate Concept Lattice and Research on its Application in Knowledge Discovery*

Aiming at the problem that the classic concept lattice is limited in incomplete information and it cannot acquire knowledge of association rules in Missing-Value Context very well, an approximate concept lattice about parameter of Tau was put forward, based on lower approximate formal context, to solve the problem of knowledge discovery extracted by association rules with incomplete information, by which the association knowledge rules with higher accuracy and credibility can be acquired. The construction of approximate concept lattice about parameter of Tau and the extraction process of association rules were described in detail, and through case analysis and comparison with two solution strategies of lower and upper approximate concept lattice, the feasibility and effectiveness of that method are verified.