Link prediction in multiplex networks: An evidence theory method

Due to its broad range of applications, link prediction has captured considerable attention from various disciplines. In this paper, we focus on the problem of link prediction in multiplex networks. As a particular case of complex networks, multiplex networks can describe the complex systems that include different types of relationships between the same group of entities. In a multiplex network, different layers are typically interrelated. In this regard, we propose a new multiplex link prediction method that gauges the connection likelihood of a node pair by integrating its similarity scores from all layers using evidence theory. In the proposed method, each layer is regarded as a source of evidence, and the similarity of a node pair in one layer is represented by a mass function. Thus, we define a new index to compute the fundamental similarity of a node pair in a layer. We also estimate the reliability of evidence via layer relevance. To analyze the effectiveness of the proposed method, we conducted extensive experiments over eight multiplex networks, and results show that the proposed method outperforms existing methods in most cases.