Time Frame based Link Prediction in Directed Citation Networks

Abstract

Link prediction is a well-known problem in field of social network analysis which intends to guess the likelihood of the occurrences of connections between nodes. By using the structure of the network up to a given time appearance of links in future can be predict. In the most of previous studies, for performing the link prediction task just according to the exploration of the state of the network at a specific moment by applying the proximity metrics like topological based metrics to non-connected nodes has been used in order to predict new links. In those studies the behavior of links along the time or directed networks didn't considered which can be point as a limitation in link prediction studies. In this study we tried to overcome the above mentioned limitation by analyzing the development of topological measures in a citation network on a specific pried of time. For achieving this aim, chosen similarity matric deployed to all non-connected pairs of nodes in different frames of time in the network. Then, time frames are built for each pair to record their values which provided by the metric. Experiments on unsupervised prediction on a directed citation network show that the proposed method finds satisfactory results and is promising.