Stock Market Prediction Based on Financial News, Text Data Mining, and Investor Sentiment Analysis

Abstract

In the financial market, the stock market, as a crucial component, attracts widespread attention. However, traditional stock market predictions mainly rely on historical data, overlooking the influence of financial news and investor sentiment on the market. This study proposes a novel stock market prediction model, CAB-LSTM, utilizing financial news text data mining and sentiment analysis. The model integrates news topics and sentiment features, demonstrating higher accuracy compared to traditional models. Research results indicate that the CAB-LSTM model effectively forecasts stock market trends, mitigating trading risks, and offering new perspectives for investment decisions. This study provides theoretical support for stock market prediction and sentiment analysis based on financial news text data mining.