[5N2-GS-11h] AI and Society

1:00 PM - 1:15 PM JST(4:00 AM - 4:15 AM UTC)

[5N2-GS-11h-05] Interaction Analysis of Influencer Posting Behavior and Engagement

〇Yuki Horikawa¹, Fujio Toriumi¹, Seiichi Igaya², Chihiro Iwai² (1. The University of Tokyo, 2. Hakuhodo DY Holdings Inc.)

Keywords:

Social Media,Computational Social Science

In recent years, the influence of social media influencers has grown significantly, raising concerns about their potential role not only in disseminating information but also in amplifying misinformation and polarized opinions. Their behavioral patterns are considered to be closely linked to user reactions (e.g., likes, retweets, comments); however, the long-term dynamic interaction mechanisms remain insufficiently understood. Prior research has identified behavioral shifts where general users curate posts with an "imagined audience" in mind or strategically withhold similar content when receiving higher-than-expected engagement. This study analyzes how influencers adapt their content, posting frequency, and self-presentation strategies in response to continuous user feedback—specifically focusing on the divergence between expected and actual engagement (positive or negative deviations). Specifically, we examine how the magnitude of these reactions influences shifts between "private" and "public" personas and the inclination toward extreme discourse. By clarifying the adaptive strategies influencers employ to gain engagement, this research provides insights aimed at preventing the spread of misinformation and fostering a healthy information environment.