AI‑enabled Investor Sentiment Analysis and Stock Market Prediction

Abstract

The efficient market hypothesis assumes that investors have complete rationality, while traditional theories advocate that stock prices are determined by the fundamentals of the enterprise. Previous studies have often used lagged trading indicators such as turnover rate and fund discount rate to measure investor sentiment. However, these methods have insufficient accuracy and poor timeliness, and cannot capture retail public opinion in real time, explain frequent stock price fluctuations and various market anomalies in the stock market. With the continuous development of natural language processing and big model technology, relying on artificial intelligence text quantification methods to analyze investor emotions contained in financial information, stock bar comments, and social platform public opinion has become a popular research direction in the fields of behavioral finance and intelligent finance. However, there are still many shortcomings in existing related research. Most existing literature separately explores one direction, namely AI text technology, investor sentiment transmission mechanism, and stock market prediction model, and has not yet systematically integrated the relationship between the three. Most studies have not comprehensively sorted out the new market risks caused by false information and emotional interference, and the overall research framework is relatively scattered. Based on this, this article provides a systematic review of the application of artificial intelligence in investor sentiment monitoring and stock market prediction, focusing on three issues: how AI technology can compensate for the shortcomings of traditional sentiment measurement methods; The transmission channel of investor sentiment on short-term stock market fluctuations; The research gap, new risks, and response strategies left by generative AI for market sentiment analysis. I hope to clarify the internal logic of artificial intelligence empowering investor sentiment analysis by sorting out the above research context, making up for the fragmented shortcomings of the existing research system, and providing reference ideas for subsequent empirical research and stock market risk prevention and control.