Feasibility of using ChatGPT to predict article retraction from Twitter mentions

Ascertain whether ChatGPT can be utilized to predict the retraction of scholarly articles based solely on the textual content of Twitter mentions referencing those articles.

Background

LLMs like ChatGPT have demonstrated strong capabilities in natural language understanding and annotation tasks, suggesting potential applications in predictive tasks based on social media text. The question is whether ChatGPT can infer retraction risk from tweets about scholarly articles.

Resolving this problem would establish whether generative AI can assist in early detection of problematic publications by processing reader comments and discussions on social media, complementing human judgment and traditional computational methods.