Approximating Network Centrality Measures Using Node Embedding and Machine Learning

Extracting information from real-world networks has become a key challenge due to the large sizes such networks achieve nowadays. Depending on the intended node centrality, it becomes unfeasible to compute it for such large complex networks due to the computational cost. One way to tackle this problem is by developing fast methods capable of approximating network centralities. In this paper, we propose an approach for efficiently approximating node centralities for large networks using Neural Networks and Graph Embedding techniques. Our proposed model, entitled Network Centrality Approximation using Graph Embedding (NCA-GE), receives the adjacency matrix of a graph and the degree of every node as input and computes the approximate desired centrality rank for every node. NCA-GE has a time complexity of O(|E|), E being the set of edges of a graph, making it suitable for large networks. NCA-GE also trains pretty fast, requiring only a set of small synthetic graphs, and it works well for different node centralities and different network sizes and topologies. Finally, we compare our approach to the state-of-the-art method that approximates centrality ranks using the degree and eigenvector centralities as input, where we show that the NCA-GE outperforms the former in a variety of scenarios.