Studying Customer Groups from Blogs

Blogs have become increasingly popular and have been widely used for such purposes as online diaries, commentaries, and socialization. In this paper we present our research on extraction of useful customer group information from blogs. We use both content analysis and structural analysis methods to identify important bloggers who either blog a lot about a product (e.g., iPod) or occupy key positions in the network structure. We also study the online communities formed by bloggers who hold different attitudes toward the product in order to see how attitudes affect their interaction patterns. Some of our preliminary findings are surprising and worth further study.