Content Analysis of Items in Newspaper Data Using Table Arrangement Technology and ChatGPT for Stock Price Prediction
Masaki Murata · 2023
In this study, we employed table arrangement techniques and ChatGPT to analyze newspaper content relevant to stock prices. Using table arrangement techniques, we effectively organized sentences from articles into tables, extracting 22 key content elements. Additionally, we discovered that ChatGPT possesses the ability to extract and present newspaper data in tabular form. Factors were found to influence stock price movements. Drops in stock prices were impacted by factors such as crude oil prices, and the COVID-19 pandemic. Conversely, rising stock prices were supported by global trends, and vac-cine effectiveness. Furthermore, we propose a highly effective, large-scale method for constructing tables by combining table arrangement techniques and ChatGPT. The proposed method achieves an accuracy rate of 0.95 under a lenient criterion.