Composite Information Integration System for Stock Price Forecasting

The main aim is to improve stock price predictions by combining deep learning techniques with financial market analysis. The difference between the other model prediction and our model prediction is it uses both stock data andnews articles to develop a robust predictive model. A composite information mixing module integrates time-series data and news text features. The model’s performance is confirmed by metrics like accuracy, precision, recall, and F1-score, which show how well it works. A novel hybrid approach is proposed, combining price and text data using deep learning models like ARIMA, LSTM, GRU, and BiLSTM to extract time-series features from stock price data and semantic features from news text data. Apply the Multilayer Perceptron (MLP) based model to make accurate predictions. This initiative demonstrates the benefits of integrating diverse data sources and considering market volatility, and gives helpful insights to financial analysts and decision-makers in the stock market.