Review on Graph Feature Learning and Feature Extraction Techniques for Link Prediction

The problem of link prediction has recently attracted considerable attention by research community. Given a graph, which is an abstraction of the relationships among entities, the task of link prediction is to anticipate future connections among entities in the graph, concerning its current state. Extensive studies have examined this problem from different aspects and proposed various methods, some of which might work very well for a specific application but not as a global solution. This work presents an extensive review of state-of-art methods and algorithms proposed on this subject and categorizes them into four main categories: similarity-based methods, probabilistic methods, relational models, and learning-based methods. Additionally, a collection of network data sets has been presented in this paper, which can be used to study link prediction. To the best of our knowledge, this survey is the first comprehensive study that considers all of the mentioned challenges and solutions for link prediction in graphs with the improvements in the recent years, including the unsupervised and supervised techniques and their evolution over the recent years.