Public Opinion Mining using Large Language Models on COVID-19 Related Tweets
Vu Tran(統計数理研究所)
Social media platforms have emerged as a significant source of public opinion, offering a massive user-generated data in which user-opinions are valuable if obtainable. Large language models (LLMs) have been in the spotlight recently with suggestions on the emergent abilities to solve tasks that are not explicitly trained for. Thus, this study explores the potential of utilizing LLMs for opinion mining on social media data by asking LLMs difficult questions, instead of simply asking whether the text’s sentiment polarity is either positive, negative, or neutral. This study compares the LLM response statistics and the corresponding public surveys related to COVID-19, including the intention to take vaccination and the stress check. The results indicate that it is promising, but also challenging, to utilize LLMs for the tasks.
On Japanese COVID-19 Policy Effect: Evidence from Tokyo and Osaka
毛柏林(京都大学経済研究所、特定助教)
In this study, within a DID framework, we employ two public mobility indices based on NTT Docomo Mobile Kukan Toukei data, and SMCC credit card consumption data to estimate the effect of Japanese COVID-19 policy (the Emergency Declaration and the Manbo Policy) on (1) public mobility, (2) offline credit card consumption by sectors and (3) the speed of infection spreading in Tokyo and Osaka. Throughout the targeted pandemic period—from the first wave to the fifth wave—we find that the policy effectively controls the public mobility and consequently reduces the speed of infection spreading. However, the policy also results in adverse effects on offline credit card consumption across most sectors. Furthermore, the policy effect is heterogeneous, with more influential effect during the first wave compared to subsequent waves. Our estimated policy effect provides evidence for future pandemic policy-making.