Optimizing Stock Market Prediction Through Integration of Market News and Stock Prices with Machine Learning Models

This study investigates the enhancement of stock market prediction accuracy by integrating market news sentiment with historical stock price data using advanced machine learning models. Traditional prediction methods primarily rely on historical price data, often neglecting the influence of real-time news events on market behavior. By using the benefits of both Long Short-Term Memory (LSTM) networks and Bidirectional Encoder Representations from Transformers (BERT), hybrid model aims to close this gap. The LSTM component finds temporal connections in stock price data, whereas BERT measures the impact of news item sentiment analysis on stock prices. A dataset of stock prices and related news stories is utilized to test the model, and performance measurements including Mean Absolute Error (MAE) and R-squared (R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>) are employed. Results demonstrate that the integrated model significantly outperforms traditional methods, particularly in scenarios where market volatility is driven by news events. This study demonstrates the potential for enhancing the precision of stock market forecasts by merging various data sources, particularly historical price data and real-time news sentiment. The insights this provides can be of great use to automated trading systems, advisors, and investors.