Analysis Using Popularity Awareness Index, Recency Index and Purchase Diversity in Group Buying

We propose new metrics for customers’ purchasing behaviors in a group buying coupon website, based on HITS algorithms and information entropy: that is, popularity awareness index, recency index, and purchase diversity. These indices are used to classify customers and predict future behaviors. This paper includes definitions of these new indices to be used in real group buying websites. In these websites, adequate characteristics for customers are strongly required and are critical for marketing purpose. We will also provide some experimental results on real data set, including customer segmentation used in future marketing planning.