Applying Link Prediction Techniques for Item Recommendation in C2C Commercial Networks

There has been a big revolution in electronic commerce since the advent of recommender systems. Most of the current recommender systems are designed for B2C e-commerce sites. But this paper focuses on building a recommendation algorithm that increases volume and speed of forming trades between users by considering special features of C2C e-commerce sites. In this paper, we consider users and transactions between them as a network in which nodes represent users and edges represent transactions between them. By this mapping, link prediction approaches could be used to build the recommender system. The proposed model, rather than topology of the network, uses nodes’ features like: category of items, ratings of users, and reputation of sellers. The results show that the proposed model can be used to predict future trades between users in a C2C commercial network.