Mining fuzzy similar sequential patterns from quantitative data

摘要

Data mining of sequential patterns from items in transaction databases has been studied extensively in recent years. In order to discover more practical rules, domain knowledge such as taxonomies of items and similarity among items have been considered to produce multiple-level sequential patterns and similar sequential patterns respectively. However, these algorithms deal with only transactions with binary values whereas transactions with quantitative values are more commonly seen in real-world applications. This paper thus proposes a new data-mining algorithm for extracting fuzzy knowledge from transactions stored as quantitative values. The proposed algorithm integrates fuzzy set concepts and the Aprioriall mining algorithm to find fuzzy similar sequential patterns in a given transaction data set where similarity relations are assumed among database items. The rules discovered here thus promote coarser granularity of sequential patterns and exhibit quantitative regularity under similarity relations. The results developed here can be applied to cross-marketing analysis, web usage mining, etc.