Character analysis of online consumers based on Bayesian network

Clickstream data provide information for the behavior traces of online consumers. These traces contain the user's personality psychology. In this paper, clickstream data is researched to analyze the online consumer character in personality psychology. We propose a consumer character analysis method by Bayesian network. By using some variables at the session level, we construct the Bayesian network and dig out the consumer character. Then, with some real clickstream data, this method is applied to an online shopper. Finally, we get the dominant type of consumer character and related variables. Through interview with user, the validity of this method is verified.