"ChatGPT应用于股票市场预测：情感分析与股价走势预测"

The paper "Can ChatGPT Forecast Stock Price Movements? Return Predictability and Large Language Models" explores the use of ChatGPT and other large language models in predicting stock market returns through sentiment analysis of news headlines. The researchers, Alejandro Lopez-Lira and Yuehua Tang from the University of Florida, implemented ChatGPT to determine whether a news headline is positive, negative, or irrelevant to the stock prices of specific firms. The study focuses on the potential of utilizing ChatGPT as a tool for stock market prediction by analyzing the sentiment of news headlines. By assigning numerical scores to the sentiment of each headline, the researchers aim to forecast stock price movements more accurately. The research, conducted in April 2023, highlights the growing interest in incorporating artificial intelligence and natural language processing technologies into financial analysis. Through the use of ChatGPT, a language model widely known for generating human-like responses in conversations, the researchers seek to enhance the efficiency and accuracy of stock market predictions. Overall, the study delves into the possibilities of leveraging ChatGPT and other large language models in financial applications by analyzing sentiment in news headlines to forecast stock price movements. The findings suggest that incorporating advanced language models in stock market analysis could potentially lead to improved prediction accuracy and decision-making in the financial sector.