CONSUMER BEHAVIOR ANALYSIS BY GRAPH MINING TECHNIQUE
Katsutoshi Yada, Hiroshi Motoda, Takashi Washio, Asuka Miyawaki · New Mathematics and Natural Computation · 2006
In this paper, we discuss how graph mining system is applied to sales transaction data so as to understand consumer behavior. First, existing research of consumer behavior analysis for sequential purchase pattern is reviewed. Then we propose to represent the complicated customer purchase behavior by a directed graph retaining temporal information in a purchase sequence and apply a graph mining technique to analyze the frequent occurring patterns. In this paper, we demonstrate through the case of healthy cooking oil analysis how graph mining technology helps us understand complex purchase behavior.