Novel Influence Maximization Algorithm for Social Network Behavior Management

Abstract

The users largely contributing towards product adoption or information utilization in social networks are identified by the process of influence maximization. The exponential growth in social networks imposes several challenges in the analyses of these networks. Important has been given to modeling structural properties while the relationship between users and their social behavior has being ignored in the existing literature. With respect to the social behavior, the influence maximization task has been parallelized in this paper. In order to maximize the influence in social networks, an interest based algorithm with parallel social action has been proposed. This is algorithm enables identifying influential users in social network. The interactive behavior of the user is weighted dynamically as social actions along with the interests of the users. These two semantic metrics are used in the proposed algorithm. An optimal influential nodes set is computed by implementing the machines with CPU architecture with perfect parallelism through community structure. This helps in reducing the execution time and overcoming the real-word social network size challenges. When compared to the existing schemes, the proposed algorithm offers improved efficiency in the calculation speed on real world networks.