A comparative study on effect of news sentiment on stock price prediction with deep learning architecture

• Published in: PloS one Vol. 18; no. 4; p. e0284695
• Main Authors: Dahal, Keshab Raj, Pokhrel, Nawa Raj, Gaire, Santosh, Mahatara, Sharad, Joshi, Rajendra P, Gupta, Ankrit, Banjade, Huta R, Joshi, Jeorge
• Format: Journal Article
• Language: English
• Published: United States Public Library of Science 25.04.2023; Public Library of Science (PLoS)
• Subjects: Accuracy; Algorithms; Analysis; Artificial Intelligence; Attitude; Benchmarking; Biology and Life Sciences; Comparative studies; Computer and Information Sciences; Correlation coefficient; Correlation coefficients; Current events; Deep Learning; Digital currencies; Dow Jones averages; Efficient markets; Industrialized nations; Influence; Internet stocks; Long short-term memory; Machine learning; Macroeconomics; Mathematical models; Neural networks; News; Normal distribution; Physical Sciences; Predictions; Prices; Prices and rates; Pricing; Reproducibility of Results; Research and Analysis Methods; Robustness (mathematics); Root-mean-square errors; Securities markets; Short term memory; Social Sciences; Statistical analysis; Statistical tests; Stock exchanges; Stocks; Trump, Donald J; Volatility; United States
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0284695

The accelerated progress in artificial intelligence encourages sophisticated deep learning methods in predicting stock prices. In the meantime, easy accessibility of the stock market in the palm of one's hand has made its behavior more fuzzy, volatile, and complex than ever. The world is looking at an accurate and reliable model that uses text and numerical data which better represents the market's highly volatile and non-linear behavior in a broader spectrum. A research gap exists in accurately predicting a target stock's closing price utilizing the combined numerical and text data. This study uses long short-term memory (LSTM) and gated recurrent unit (GRU) to predict the stock price using stock features alone and incorporating financial news data in conjunction with stock features. The comparative study carried out under identical conditions dispassionately evaluates the importance of incorporating financial news in stock price prediction. Our experiment concludes that incorporating financial news data produces better prediction accuracy than using the stock fundamental features alone. The performances of the model architecture are compared using the standard assessment metrics -Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), and Correlation Coefficient (R). Furthermore, statistical tests are conducted to further verify the models' robustness and reliability.