With the increasing use of social media, online self-organized relief has become a crucial aspect of crisis management during public health emergencies, leading to the emergence of online self-organizations. This study employed the BERT model to classify the replies of Weibo users and used K-means clustering to summarize the patterns of self-organized groups and communities. We then combined the findings from pattern discovery and documents from online relief networks to analyze the core components and mechanisms of online self-organizations. Our findings indicate the following: (1) The composition of online self-organized groups follows Pareto’s law. (2) Online self-organized communities are mainly composed of sparse and small groups with loose connections, and bot accounts can automatically identify those in need and provide them with helpful information and resources. (3) The core components of the mechanism of online self-organized rescue groups include the initial gathering of groups, the formation of key groups, the generation of collective action, and the establishment of organizational norms. This study suggests that social media can establish an authentication mechanism for online self-organizations, and that authorities should encourage online interactive live streams about public health issues. However, it is important to note that self-organizations are not a panacea for all issues during public health emergencies.

The Sci-Tech Commissioner System (SCS) is a result of exploratory efforts by the Chinese government to use science and technology to strengthen the agricultural sector. Social network analysis (SNA) and machine learning (ML) techniques make it feasible to assess the service performance in China's SCS by using indicators such as group types and structure features. In this study, SNA and a clustering algorithm were employed to categorize service group types of sci-tech commissioners. By comparing the accuracy of different classification algorithms in predicting the clustering results, LightGBM algorithm was finally select to determine the clustering features of sci-tech commissioners and establish an interpretable ML model. Then, the SHAP was used to algorithm to analyze influences affecting service performance. Results show that the service forms of sci-tech commissioners are group-oriented, and that group types include small groups of young commissioners with close cooperation, larger groups of young and middle-aged commissioners, small groups of middle-aged and old commissioners with close cooperation, and isolated points of highly-influential commissioners. Furthermore, while group size is not the determinant of a commissioner's average performance, group structure and coordination ability were found to be more critical. Moreover, while differences in distinct types of service performance are caused by various factors, but good group structures and extensive social contacts are essential for high service performance.