Abstract:Recently, link prediction has attracted more attentions from various disciplines such as computer science, bioinformatics and economics. In this problem, unknown links between nodes are discovered based on numerous information such as network topology, profile information and user generated contents. Most of the previous researchers have focused on the structural features of the networks. While the recent researches indicate that contextual information can change the network topology. Although, there are number of valuable researches which combine structural and content information, but they face with the scalability issue due to feature engineering. Because, majority of the extracted features are obtained by a supervised or semi supervised algorithm. Moreover, the existing features are not general enough to indicate good performance on different networks with heterogeneous structures. Besides, most of the previous researches are presented for undirected and unweighted networks. In this paper, a novel link prediction framework called "DeepLink" is presented based on deep learning techniques. In contrast to the previous researches which fail to automatically extract best features for the link prediction, deep learning reduces the manual feature engineering. In this framework, both the structural and content information of the nodes are employed. The framework can use different structural feature vectors, which are prepared by various link prediction methods. It considers all proximity orders that are presented in a network during the structural feature learning. We have evaluated the performance of DeepLink on two real social network datasets including Telegram and irBlogs. On both datasets, the proposed framework outperforms several structural and hybrid approaches for link prediction problem.

Link Prediction Via Deep Learning

Link Prediction via Graph Attention Network

Link Prediction through Deep Generative Model

Link Prediction Based on Deep Convolutional Neural Network

Generative Graph Neural Networks for Link Prediction

DeepLink: A Novel Link Prediction Framework based on Deep Learning

Link prediction based on depth structure in social networks

Link Prediction Via Linear Optimization

Virtual-Link Representation for Link Prediction

Learning Universal Network Representation Via Link Prediction by Graph Convolutional Neural Network

Deep Non-Negative Matrix Factorization with Edge Generator for Link Prediction in Complex Networks

Network Embedding with Enhanced Feature Representations for Link Prediction.

Link Prediction in Social Networks Based on Hypergraph

A Directed Link Prediction Method Using Graph Convolutional Network Based on Social Ranking Theory

Deep Dynamic Network Embedding for Link Prediction

A Method for Predicting Links in Complex Networks by Integrating Enclosure Subgraphs With High-Frequency Graph Information

Line Graph Neural Networks for Link Weight Prediction

A Link Prediction Method for Large-Scale Networks

Deep learning approaches for link prediction in social network services

A Representation Learning Link Prediction Approach Using Line Graph Neural Networks.

Dynamic link prediction by learning the representation of node-pair via graph neural networks