Mining Buyer Behavior Patterns Based on Dynamic Group-Buying Network

Abstract

New challenges arise as analysts started to apply traditional network analysis techniques to social network in the Web 2.0 era. Many business modes in e-commerce emerged based on online or offline social network, such as group-buying, social buying and viral marketing. The target environments are those involving large amounts of relational data that is time-dependent. Such characteristics require time-sensitive network analyses to support decision making. For example, in the dynamic online markets, understanding changes in buyer behavior and the dynamics of social networks can help manager to establish effective promotion campaigns. However, traditional single snapshot approach does not exactly fit in the time-sensitive network representation scenarios. Thus, this chapter first proposed a time-sensitive network by using timestamp to enhance edge representation, and then provided a methodology based on the framework of business intelligence platform to support dynamic network modeling, analysis and data mining. Finally, China’s case study described the process of dynamic group-buying network modeling and analysis, as well as the results of the buyer behavior pattern analysis and mining.