Link prediction algorithm for co-authorship networks using Neural Network

Link prediction problem may be applied in many different domains. In the network, nodes represent authors and a connection between them indicates they coauthored a paper. This paper describes an improved method for predicting possibility of a link between authors. The algorithm is implemented in C# programming language and testing has been performed on Intel Core2Duo processor with 4 GB RAM and 2 GHz processor speed. It was observed that this method shows better performance compared to the previously presented methods.