Research on Weibo Rumor Propagation Mechanism Based on Big Data

Abstract

Weibo's wide reach and openness make it a platform for both influence and the spread of rumors, causing negative social impacts. Therefore, studying the dissemination of rumors on Weibo holds important practical significance. This research takes the recent "Starch Sausage Incident" as a case study, expanding the traditional SIR model by incorporating a rational population, resulting in the SIRW model. Utilizing web scraping techniques to collect Weibo data, the study conducted data preprocessing and performed sentiment analysis using a GRU model. By combining sentiment results with user profile information, the research analyzed the cumulative numbers in various compartments over different time periods. Parameters were fitted using the least squares method, and the results indicate that the model performs well. Considering the characteristics of Weibo, the study employed a BA scale-free complex network for visual simulation and analyzed the impact of the degree and number of initial infected nodes on rumor propagation. Through robustness analysis, the research explored effective strategies for interrupting rumor transmission from the perspectives of large nodes and bridge nodes, proposing targeted public opinion control recommendations. The study demonstrates that a higher initial proportion of rational individuals significantly restricts rumor spread; conversely, a moderate number of rumors may enhance individuals' rationality and discernment. During different stages of rumor propagation, targeting hub nodes or bridge nodes can markedly improve the effectiveness of disruption.