关联规则格的渐进式维护算法

Incremental Algorithm for Maintaining Association Rule Lattice

关联规则发现作为数据挖掘中核心任务之一,已经得到了广泛的研究。由二元关系导出的概念格是一种非常有用的形式化工具,适于发现数据中潜在的概念。在分析了概念格和关联规则之间的关系的基础上,根据需要对概念格结构——关联规则格进行了修改,同时,采用了带头尾指针的链表作为整体的数据结构,从而提出了基于关联规则格的关联规则渐进式维护算法。该算法可以根据预先给定的置信度∮,在渐进式增加和删除节点时,动态更新关联规则。

Association rule discovery, as a kernel task of data mining, has been studied widely. And concept lattice, induced from a binary relation between objects and features, is a very useful formal tool and is fit for discovering the potential concept below the data. The relationship between concept lattice and association rule discovery is analyzed. The structure of node in lattice, which is called association rule lattice is modified according to the requirement and the data structure of linked-list with a head finger and a trail finger is adopted. Based on this, the incremental algorithm for maintaining association rule is put forward. According to ∮ ,which is the confidence, the algorithm can be re^eshing the association rules dynamically when adding and deleting object one by one.