Who tweets about quantum physics research on Twitter: the impact of user types, tweet content and interaction patterns

Understanding the use and users of social media in the context of scientific communication is essential. Analyzing and identifying different types of public engagement provides valuable insights into how academics and broader society interact with scientific outputs. This study takes quantum physics papers as samples and employs machine learning, scientometric methods, and regression analysis to unravel the interaction patterns and characteristics of various users. The user identification results indicate that, in addition to academics, several professional groups actively participate in disseminating scientific papers on Twitter. Using the density peak clustering method, the study profiles users and finds that those who play key roles in the dissemination network exhibit distinct interactive features in their social media behavior. Examining interaction patterns, including mentions, retweets, and quotes, reveals that users display homogeneity in the mentioning pattern, while in retweets and quotes, they are more inclined to engage with the academic category. Furthermore, the sentiment orientation, media richness, and dialogic loop elements of scientific tweets have varying influences on user engagement, with user types serving as significant moderators. By integrating Altmetrics metrics, this study contributes to the literature by providing empirical evidence that the involvement of diverse stakeholders enhances the dissemination of academic achievements on social media. These findings improve our understanding of the societal impact of publications and shed light on the interactions between science and society.