HCT - A Hybrid Algorithm for Influence Maximization Problem Based on Community Detection and TOPSIS.

TL;DR: This paper proposes HCT, a hybrid algorithm for influence maximization in social networks, combining community detection and TOPSIS to evaluate node influence, achieving better accuracy and efficiency than existing algorithms on six real-world networks.

Abstract: The influence maximization problem is to find a subset of nodes in the social networks for the purpose of maximizing the number of nodes that the subset of nodes can influence. The influence maximization problem is an open issue in the analysis of the social networks. Many algorithms have been proposed to solve this problem. However, most existing algorithms usually do not have an acceptable accuracy or efficiency. Therefore, this paper proposes a new algorithm as a tradeoff between the accuracy and efficiency, called A Hybrid Algorithm Based on Community Detection and TOPSIS (HCT). The HCT algorithm proposes two new metrics based on the community detection to evaluate the influence of a node, called Direct Influence between Communities (BDS), Indirect Influence between Communities (BIDS), respectively. Moreover, The Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) is used to identify the most influential nodes. Furthermore, the entropy weight method is used to overcome the shortcoming of the TOPSIS method, which can also improve the accuracy of the proposed algorithm. The experimental results on six realworld networks show the proposed algorithm have a better accuracy and efficiency than the comparison algorithms.