A Novel Stock Index Prediction Method Based on Financial Text Sentiment Analysis

With the rapid growth of artificial intelligence and machine learning, predicting stock market trends has become an increasingly important and complex task. In this paper, we propose a novel stock index prediction model, SA-BERT-LSTM, based on financial text sentiment analysis. This model integrates the BERT (Bidirectional Encoder Representations from Transformers) sentiment analysis method with the Long Short-Term Memory (LSTM) network, combining the extracted sentiment features from financial news with stock trading data. We apply this model to predict the rise and fall of the CSI 300 Index. The results demonstrate that sentiment features significantly enhance prediction accuracy. Compared to three benchmark modelsBP neural network, Support Vector Machines (SVM), and XGBoostthe SA-BERT-LSTM model outperforms them in terms of prediction precision. Furthermore, the model is adaptable for predicting individual stock prices. This research contributes to the application of sentiment analysis in financial market prediction, broadening the scope of AI and machine learning in economics, and provides valuable insights for the implementation of national AI strategies.