Link Prediction In Multi-layer Collaboration Network

The prediction of relationships in a social network is a complex and extremely useful task to enhance or maximize collaborations by indicating the most promising partnerships.Thus,predicting/recommending future links in social networks has attracted a great deal of attention.Link prediction has many applications and,it offers many benefits to the users of social networking services such as providing fast and accurate recommendations or suggestions to the users.However,highly structured massive real-world networks involving heterogeneous entities with complex associations have added new challenges to link prediction research.Recent studies have shown that real world networks are heterogeneous in nature with multiple types of links between same node pairs.Organizing these networks in a layered structure,where each layer represents a different network,can reveal interesting cross-layer interaction patterns.In coevolving networks,links in one layer result in an increased probability of other types of links/collaborations forming between the same node pair.There have been numerous attempts to address the problem of link prediction through diverse approaches.Most of the classical link prediction approaches gather single layer links information for prediction task.However,in this research work,we have presented link prediction problem for multi-layer networks.To predict links in one layer using information from different layers of a multi-layer network,we have calculated multiplex attributes from our network by extending simple node-neighborhood similarity indices in multiplex context.We have explored multiplex relations in a co-authorship network to predict future collaborations(co-authorship links)among authors.The applied approach is a supervised machine learning approach.While such an approach has been successfully applied in the context on simple networks,we have exploited this approach for a multi-layer network in which link existence likelihoods for the target layer are learned from the attributes calculated on target layer and other layers of that multi-layer network.We have showed our results from experiments conducted on real dataset extracted from the famous bibliographical database;APS(American Physical Society)that has been enriched with co-authorship information.