GPT-4 as a Twitter Data Annotator: Unraveling Its Performance on a Stance Classification Task

<p>This study utilizes a new labeled dataset of Twitter posts discussing the stance classification regarding abortion legalization in the USA. The main aim of the study is to evaluate the GPT-4 model's capability as a social media text annotator. We assess the performance of three prompt-based GPT-4 approaches using a sample of the dataset that exhibits perfect agreement among human annotators. The tweets are categorized into three labels: 'favor,' 'against,' or 'none,' representing their respective stances on the issue.</p>