Abstract: This study investigates the key factors driving rumor diffusion, traces their intertwined evolution, and analyzes the diffusion mechanisms to support listed companies in formulating effective crisis management strategies to mitigate potential losses. First, this study examines diffusion paths to identify key elements in the spread of negative rumors on social networks. Using the "Haitian double standards" incident as a case study, this research collects data from Sina Weibo and analyzes it within a life-cycle theoretical framework, employing methods such as social network analysis and LDA topic modeling. Tools like Gephi and Python are further utilized to identify key elements and trace the evolution of the diffusion network. Finally, practical crisis management strategies are proposed based on the diffusion mechanism. This study found that the three key elements in the spread of negative rumors on social networks, namely users, information content and time periods, are intertwined and evolving during the spreading process. As time progresses, user degree centrality shows an inverted U-shape. Key participants transition from grassroots, non-verified users to authoritative, verified users, and eventually back to non-verified users. Meanwhile, content themes evolve towards core issues and then transition to long-term performance, with the growth phase being a crucial period for the progression of the event. Listed companies can enhance their crisis response by targeting influential users and addressing critical content themes during this phase to proactively mitigate risks.

Key words: Information diffusion, Negative rumors, Social network analysis, Text mining, Internet public opinion