- 20 November 2026
- book chapter
- Published by IGI Global Scientific Publishing
- p. 73-102
Abstract
This chapter presents Multi Source Attention Long Short Term Memory (MSA-LSTM), a deep learning framework for real time financial forecasting that fuses three heterogeneous data streams historical OHLCV price data, financial news sentiment, and macroeconomic indicators within an Attention enhanced Long Short Term Memory (LSTM) network. A sliding window procedure converts these streams into synchronized temporal sequences, and the attention layer dynamically re weights time steps and features so the model can be inspected as well as used for prediction. The model performance is approximately 73% and 39% better than an ARIMA baseline and attention free LSTM respectively. Based on these results and an ablation analysis presented later in the chapter, as well as a statistical significance assessment performed later in the chapter, we can conclude that sentiment and macro economic context and attention based temporal modeling can be combined in a way that results in a measurable and reproducible improvement over single-sources baselines.Keywords
This publication has 12 references indexed in Scilit:
- Deep learning and machine learning models for portfolio optimization: Enhancing return prediction with stock clusteringResults in Engineering, 2025
- Enhancing stock market predictions for classifying unlabelled celebrities' twitter dataJournal of Open Innovation: Technology, Market, and Complexity, 2025
- Stock market prediction with optimized PLSTM-AL in smart urban citiesFinance Research Open, 2025
- A multi-feature selection fused with investor sentiment for stock price predictionExpert Systems with Applications, 2025
- Relational Stock Selection via Probabilistic State Space LearningIEEE Transactions on Knowledge and Data Engineering, 2024
- Forecasting Turning Points in Stock Price by Integrating Chart Similarity and MultipersistenceIEEE Transactions on Knowledge and Data Engineering, 2024
- Spatial–Spectral–Temporal Deep Regression Model With Convolutional Long Short-Term Memory and Transformer for the Large-Area Mapping of Mangrove Canopy Height by Using Sentinel-1 and Sentinel-2 DataIEEE Transactions on Geoscience and Remote Sensing, 2024
- Federated Learning-Inspired User Personality Prediction Using Sentiment Analysis and Topic PreferenceIEEE Transactions on Consumer Electronics, 2023
- Learning to Understand the Vague Graph for Stock Prediction With Momentum SpilloversIEEE Transactions on Knowledge and Data Engineering, 2023
- Temporal Fusion Transformers for interpretable multi-horizon time series forecastingInternational Journal of Forecasting, 2021