Abstract:

To guide stock market-related decisions, we developed a synchronized framework that employs Natural Language Processing techniques to automatically analyze public information from social media. In this project, tweets are aligned with stock market data, sentiment analysis is performed using OpenAI’s model, and market guidance is generated by applying classification models to real-time information. The models were trained using approximately 188,000 tweets and three major indices from the New York Stock Exchange for the year 2021, and tested with 111,000 tweets posted in May 2024. The models demonstrated consistently strong performance. This framework can be extended to analyze text documents from various sources, given appropriate training datasets. Index Terms—Sentimental Analysis, Social Media, Stock Market Change, Twitter, Classification, Random Forest Classifier.