Differences in Behavioral Characteristics and Diffusion Mechanisms: A Comparative Analysis Based on Social Bots and Human Users

In the field of social media, the systematic impact that bot users bring to the dissemination of public opinion has been a key concern of the research. To achieve more effective opinion management, it is important to understand how and why behavior differs between bot users and human users. The study compares the differences in behavioral characteristics and diffusion mechanisms between bot users and human users during public opinion dissemination, using public health emergencies as the research target, and further provides specific explanations for the differences. First, the study classified users with bot characteristics and human users by establishing the relevant formulas of user indicator characteristics. Secondly, the study used deep learning methods such as Top2Vec and BERT to extract topics and sentiments, and used social network analysis methods to construct network graphs and compare network attribute features. Finally, the study further compared the differences in information dissemination between posts published by bot users and human users through multi-factor ANOVA. It was found that there were significant differences in behavioral characteristics and diffusion mechanisms between bot users and human users. The findings can help guide the public to pay attention to topic shifting and promote the diffusion of positive emotions in social networks, which in turn can better achieve emergency management of emergencies and the maintenance of online orders.