People to People Recommendation using Coupled Nonnegative Boolean Matrix Factorization
Thirunavukarasu Balasubramaniam, Nayak Richi, Chau Yuen · 2018
In the era of Web 3.0, people to people recommendation is important to identify and suggest the potential person who one might be interested to connect with. In most of the cases, the interaction between people generated is very sparse as people usually connect within a fewer circle. Matrix factorization has been successfully employed in generating recommendations of items to users under sparse condition and usually user to user friendship is used as an additional trust information in generating more accurate item recommendations. In this paper we develop a coupled matrix factorization model to accurately generate people to people recommendation by utilizing users' interaction behaviour with items. Our empirical results shows that the proposed model achieves higher quality than the state of the art methods.