Community-based influence maximization in social networks under a competitive linear threshold model considering positive and negative user views

Social networks are a collection of people or groups and interactions between them, which can include friendship activities or business relationships, and they play an important role in spreading information and maximizing influence on users. The effect of social networks in our daily life is undeniable and it shows new methods of communication and is used as a medium to spread news, opinions, thoughts and any other kind of information. The main goal in influence maximization is to find a subset of key users that maximize influence expansion under a particular diffusion model. A number of studies over the past several years have tried to solve this problem by considering a nonracist environment where there is only one player without competitors, whereas in the real world, there is always more than one player competing with other players. To influence most nodes, this problem is called competitive influence maximization (CIM). Therefore, we try to solve the problem of maximizing the competitive influence by proposing a new diffusion model, considering two positive and negative views towards advertising, which is an extension of the linear threshold (LT) model with the ability to decide on the expansion of input influence and leads to an improvement in selecting seeds with better quality and also by extracting the community structure from the input graph, it is used to calculate the expansion of each node locally within its own community, and by preventing the repetition of penetration calculations for each node, it significantly reduces the execution time. The purpose of the proposed algorithm is to find the minimum number of seed nodes that can achieve more expansion compared to the expansion of nodes selected by other competitors in an acceptable execution time. Real-world experiments and synthetic datasets show that the proposed approach can obtain the minimum node required to defeat the competitor more realistically and in less time. Competitor leads to budget cuts by reducing the number of seeds required to defeat. The obtained results show that if n seed nodes are needed to defeat the competitor, the execution time of the proposed algorithm will be approximately 1/n.