Minimizing the influence of dynamic rumors based on community structure. Issue 3 (2nd September 2019)

• Record Type: Journal Article
• Authors: Wu, Qingqing; Zhao, Xianguan; Zhou, Lihua; Wang, Yao; Yang, Yudi
• Is Part Of: International journal of crowd science. Volume 3:Issue 3(2019)
• Page Start: 303
• Page End: 314
• Publication Date: 2019-09-02
• Subjects: Social network -- Community structure -- Influence propagation -- Minimization of influence -- Rumor blocking; Human-computer interaction -- Periodicals; Human computation -- Periodicals; Cooperating objects (Computer systems) -- Periodicals; 621.3984
• DOI: 10.1108/IJCS-09-2019-0025
• Languages: English
• ISSNs: 2398-7294

Abstract

Abstract : Purpose: With the rapid development of internet technology, open online social networks provide a broader platform for information spreading. While dissemination of information provides convenience for life, it also brings many problems such as security risks and public opinion orientation. Various negative, malicious and false information spread across regions, which seriously affect social harmony and national security. Therefore, this paper aims to minimize negative information such as online rumors that has attracted extensive attention. The most existing algorithms for blocking rumors have prevented the spread of rumors to some extent, but these algorithms are designed based on entire social networks, mainly focusing on the microstructure of the network, i.e. the pairwise relationship or similarity between nodes. The blocking effect of these algorithms may be unsatisfactory in some networks because of the sparse data in the microstructure. Design/methodology/approach: An algorithm for minimizing the influence of dynamic rumor based on community structure is proposed in this paper. The algorithm first divides the network into communities, and integrates the influence of each node within communities and rumor influence probability to measure the influence of each node in the entire network, and then selects key nodes and bridge nodes in communities as blocked nodes. After that, a dynamic blocking strategy is adopted to improve the blocking effect of rumors. Findings: Community structure is one of the most prominent features of networks. It reveals the organizational structure and functional components of a network from a mesoscopic level. The utilization of community structure can provide effective and rich information to solve the problem of data sparsity in the microstructure, thus effectively improve the blocking effect. Extensive experiments on two real-world data sets have validated that the proposed algorithm has superior performance than the baseline algorithms. Originality/value: As an important research direction of social network analysis, rumor minimization has a profound effect on the harmony and stability of society and the development of social media. However, because the rumor spread has the characteristics of multiple propagation paths, fast propagation speed, wide propagation area and time-varying, it is a huge challenge to improve the effectiveness of the rumor blocking algorithm.