Rumor Suppression Maximization Algorithm Based on Local Dominance Tree
Xu Jingke, Di Zhao, Lei Zhou · 2025
With the advancement of computer technology and the convenience of information dissemination, online social networks have become a significant platform for the spread of various types of negative information. Rumors, as a prominent form of negative information on social media, can cause social unrest and lead to economic losses. Therefore, how to quickly and effectively curb the spread of rumors has become a key issue in current social network research. Based on the Independent Cascade Model, combined with graph sampling techniques and the Three Degrees of Influence principle, an improved rumor suppression maximization algorithm, RSM-LDT , is proposed. Firstly, considering the role of deep network node structures in influence diffusion, a local dominance tree is defined to calculate the potential influence of nodes. Secondly, based on the Independent Cascade Model, the node with the highest potential influence is selected as the blocking node each time. The effectiveness of the RSM-LDT algorithm is verified by comparing its rumor suppression effect and running time with five existing algorithms on two publicly available real datasets. Experimental results show that the blocking seed set selected by the RSM-LDT algorithm proposed in this paper is more accurate, significantly reducing the spread of rumors compared to heuristic algorithms, and the running time of the algorithm is four orders of magnitude faster than that of greedy algorithms. The RSM-LDT algorithm can solve the problem of rumor suppression maximization in social networks, avoiding the high time complexity of greedy algorithms while addressing the poor suppression effect of heuristic algorithms.