[2P5-GS-11] AI and Society

[2P5-GS-11-02] Analysis of Social Media Reaction to Prime Ministers' Change of Japan

〇Kunihiro Miyazaki^5,1, Taichi Murayama², Akira Matsui³, Takayuki Uchiba⁴, Haewoon Kwak⁵, Jisun An⁵ (1. The University of Tokyo, 2. University of Osaka, 3. Yokohama National University, 4. Sugakubunka Co., Ltd., 5. Indiana University)

[[Online]]

Keywords:Social media sentiment, Prime Minister's approval rate, Regression discontinuity design, Honeymoon effect, Twitter

We investigate Twitter users' reactions to the changes of the prime minister of Japan. In particular, we examine how users' sentiments changed before and after the changes. We also analyze the topics and user transitions related to the changes. The results show that the public's sentiment toward the prime minister improves when a new prime minister takes office, that sentiment changes differently for each topic, and that the topics users expect differ for each prime minister. We also found that the user base who mentioned the prime minister before and after the change was significantly replaced. These results would be helpful in understanding the public's reaction to the change of prime ministers in Japan and for effective administrative management.