A Community-Based Method for Identifying Influential Nodes Using Network Embedding

• Published in: International Journal of Information and Communication Technology Research (Tehran.Print) Vol. 14; no. 1; pp. 38 - 47
• Main Authors: Vafaei, Narges, Keyvanpour, Mohammad Reza
• Format: Journal Article
• Language: English
• Published: Iran Telecom Research Center 01.03.2022
• Subjects: influence analysis; influence maximization problem; influential nodes; social network mining
• ISSN: 2251-6107; 2783-4425
• DOI: 10.52547/itrc.14.1.38

People's influence on their friends' personal opinions and decisions is an essential feature of social networks. Due to this, many businesses use social media to convince a small number of users in order to increase awareness and ultimately maximize sales to the maximum number of users. This issue is typically expressed as the influence maximization problem. This paper will identify the most influential nodes in the social network during two phases. In the first phase, we offer a community detection approach based on the Node2Vec method to detect the potential communities. In the second phase, larger communities are chosen as candidate communities, and then the heuristicbased measurement approach is utilized to identify influential nodes within candidate communities. Evaluations of the proposed method on three real datasets demonstrate the superiority of this method over other compared methods.