关键词: 形式背景, 结构化属性, 频繁概念, 关联规则

Abstract: An algorithm with a filter function which exploring frequent concepts in a context was proposed with structured attributes in order to learn association rules. Simply changing the function, that algorithm could compute various kinds of conceptbased rules. The advantage of the presented method is to avoid the redundancy in the intent of the computed frequent concepts with the help of structured attributes. The resulted rules are, therefore, more concise and practical.

Key words: formal context, Structured attribute, Frequent concept, association rules