Social Commerce Hybrid Product Recommender

Abstract

vision for Web 3.0 (also known as Semantic Web) is the ability to create meaning out of huge quantity of qualitative data. Existing data can be interconnected for further uses. Web 2.0 focused on the users interaction with others whereas Web 3.0 focus more on the users themselves. The advantages of Semantic Web and E-commerce give rise to commerce (also referred as f-commerce). The future of business lies on the social factor and it is this factor which gives rise to a new kind of connected consumers who are becoming influential in their own right. This paper explores a very specific instance of Semantic Web - Social Recommender System. This paper discusses the likelihood of converting data into quantitative information and using this information to power recommendations. This paper first outlines the benefits of commerce over e- commerce platform. Then the related literature work regarding hybrid recommenders is discussed. Next it is discussed how to predict ratings from a user-item rating network and friend's network and then how to unify similarity matrices obtained from different networks. And lastly this paper covers the hybrid product recommender algorithm and its experimental evaluations to predict its efficiency. Keywordscommerce, hybrid product recommender, f-commerce, homophily, Cosine similarity, Smith Waterman string similarity measure, unipartite graph, bipartite graph