User interaction behaviors on social networking sites (SNSs) have led to different comment groups. We retrieved relevant literature from the Web of Science and analyzed the trends over time and co-occurrence frequencies using CiteSpace. Our findings indicate that existing research mainly focuses on horizontal community differentiation and vertical leader differentiation, but lacks a combined horizontal and vertical approach to expose the mutual relationships between the differentiations from the anlges of communities and leaders. Therefore, we proposed an analytical framework of hierarchical communities to explore the interactive influences between different communities, levels, and user identities on SNSs. The reliability and effective-ness of the proposed hierarchical community framework are validated using user interaction data from the hot topic ‘‘Two dogs biting a girl’’ on Weibo.