Bartal, Alon, and Gilad Ravid. "Member behavior in dynamic online communities: Role affiliation frequency model." IEEE Transactions on Knowledge and Data Engineering 32.9 (2019): 1773-1784.

This study introduces a novel Temporal Role-Affiliation Frequency Model (RAFM) aimed at understanding the dynamic behaviors of individuals within online social networks (OSNs). By tracking how the roles of community members evolve over time and analyzing their frequency of affiliation to these roles, RAFM can uncover latent behavioral patterns and identify influential members within the community. The application of RAFM to interaction data from four online communities demonstrated its effectiveness in revealing not only the key influencers but also members with similar behavioral patterns across different OSNs, provided these networks share compatible temporal interaction rhythms. These findings offer new insights into member behavior in OSNs through the lens of role transitions and affiliations, contributing to the broader research on online social networks.