Frequent Pattern Mining
Guandong Xu, Yu Zong, Zhenglu Yang · 2013
Frequent pattern mining is one of the most fundamental research issues in data mining, which aims to mine useful information from huge volume of data. The purpose of searching such frequent patterns (i.e., association rules) is to explore the historical supermarket transaction data, which is indeed to discover the customer behavior based on the purchased items. Association rules present the fact that how frequently items are bought together. For example, an association rule “beer->diaper (75%)” indicates that 75% of the customers that bought beer also bought diaper. Such rules can be used to make prediction and recommendation for customers and store layout. Stemmed from the basic itemset data, rule discovery on more general and complex data (i.e., sequence, tree, graph) has been thoroughly explored in the past decade. In this chapter, we introduce the basic techniques of frequent pattern mining on different type of data, i.e., itemset, sequence, tree, and graph.