An Investigation of Domain-based Social Influence on ChatGPT-Related Twitter Data

Recently, the word ChatGPT has made its way into the vocabulary of various fields that are important to society, including healthcare, education, governance, and robotics. This study examines the various reasons for social media influences using the Social Influence Theory. The study used tweets connected to ChatGPT collected from January 1st, 2023 to March 31st, 2023, totaling 402, 965 tweets. Initially, the key domains of ChatGPT discussions were identified along with the variations in the sentiments over the period. Subsequently drawing the literature related to Social Influence Theory, this study explored the relationship between subjective norm, social identity, user commitment, and features of the tweet content with social influence on the readership of a ChatGPT-related tweet. Subjective norm, identity, sentiment and domain are significant factors for social influence. The user commitment suggests a negative relationship with the social influence. Finally, the theoretical and practical implications are explained