Colonial Competition Algorithm for Predicting Information Diffusion in Social Networks

In the modern world, social network analysis has various applications including social network management, market trend analysis, identifying influential individuals, supporters, and enhancing the performance of descriptive systems. One of the existing issues in the field of information diffusion is examining the extent of dissemination in such systems. In large-scale networks, computational complexity has increased, and one of the methods to cope with this problem is using evolutionary algorithms. In this paper, a new method is proposed to increase the accuracy of predicting information diffusion in social networks by combining the colonial competition algorithm and the centrality algorithm. In the proposed method, the centrality of nodes is first calculated for all nodes in the network. Then, based on the distance between users in the social network and the speed of information dissemination among users using the colonial competition algorithm, the optimal path for dissemination is determined, and information diffusion prediction is made accordingly. In the colonial competition algorithm, the optimal path for information dissemination in the network is identified based on how the colonies move towards the colonizer. In this paper, four different datasets were examined. The results show that the proposed method leads to better results. The algorithm's performance is evaluated based on accuracy, NMI, and error criteria, showing desirable results.