8 Leveraging ChatGPT and table arrangement techniques in advanced newspaper content analysis for stock insights

This study uses table arrangement techniques and ChatGPT to analyze articles from the Mainichi Shimbun for stock price prediction. From the analysis, we identified 22 primary factors that influence the Nikkei Stock Average. We also discovered that ChatGPT could extract and present newspaper data in a tabular format. These factors significantly impact stock price fluctuations. The Nikkei Stock Average tends to rise with improved US trade relations, a strong domestic economy, and events such as the Olympics. Conversely, it tends to decline during global stock market crashes, trade tensions, and pandemics. Moreover, we propose a highly efficient, large-scale method for creating tables by integrating table arrangement techniques with ChatGPT. Using a lenient criterion, the proposed method attains an accuracy rate of 0.88. Items frequently mentioned in articles concerning the Nikkei Stock Average are systematically presented in a table, illustrating how the index rises or falls in response to these items. This table also delves deeper into the effects of exchange rate changes on the Nikkei Stock Average. Our findings offer valuable insights into the movement of the Nikkei Stock Average. Future research will further refine these techniques to improve stock prediction accuracy.