A Link Prediction Model using Similarity and Centrality based Features

With the recent development of internet, many users have started interacting and creating their relationship in online social network. These relationship and interaction between users can be represented as links in social network graph. At present, most of the link prediction methods used similarity based features only and few have used centrality based features for prediction. In this paper, we develop a link prediction model by combining similarity and centrality-based features of social network to improve the accuracy. The experimental result shows that by combing similarity and centrality based features we can improve the prediction accuracy of link prediction model.