Discover Maximum Descriptive User Groups on the Social Web

Abstract. Product reviews in E-commerce websites such as restaurants, movies, E-commerce products, etc., are essential resources for consumers to make purchasing decisions on various items. In this paper, we model discovering groups with maximum descriptively from E-commerce website of the form < i,u, s >, where i ∈ I (the set of items or products), u ∈ U (the set of users) and s is the integer rating that user u has assigned to the item i. Labeled groups from user attributes are found by solving an optimization problem. The performance of the approach is examined by some experiments on real data-sets. Keywords.