Research on online social network information diffusion detection node selection algorithm based on the random walk model

Information diffusion detection is defined as a method for choosing the most efficient observation nodes for detecting the spread of information in a given social network. It has a great significance for opinion leader mining, rumor detection, public opinion monitoring, applications, and other aspects. Information diffusion detection is difficult because it needs to consider not only the relationship structure but also the interaction structure in a social network. In this paper, the features and classification of the networking architecture and the interaction structure are analyzed. A detection node selection algorithm called the Structure Change and Diffusion Ability Rank (SCDA Rank) algorithm, based on the random walk model, is then put forward, which not only considers the structural network changes of the nodes, but also its diffusion capabilities. Experimental results show that the proposed SCDA Rank algorithm achieves satisfactory results in three targets, i.e., the coverage ratio, the hitting time, and a reduction of the infected population, compared with other similar algorithms in the Enron dataset and for real data from the Sina microblog.