AI Based Stock Market Movement Via Sentiment Analysis and Prediction

@article{1724466,
      author = {P. Vignesh, M. Srinidhi, M. Pavan, P. Surya; P. Ramu, B. Priyaranjani; M. Laxman},
      title = {AI Based Stock Market Movement Via Sentiment Analysis and Prediction},
      journal = {Iconic Research And Engineering Journals},
      year = {2026},
      volume = {10},
      number = {4},
      pages = {4780-4813},
      issn = {2456-8880},
      url = {https://www.irejournals.com/formatedpaper/1724466.pdf},
      abstract = {The stock market is a highly dynamic and complex financial system where prices fluctuate due to various economic, political, and psychological factors. Predicting stock prices accurately is a challenging task for investors and financial analysts. Traditional methods of stock analysis rely on manual interpretation, historical trends, and technical indicators, which are often time-consuming, less accurate, and influenced by human bias. Moreover, these methods do not effectively incorporate real-time data or market sentiment, which are critical for making informed decisions. To overcome these limitations, the application of Artificial Intelligence (AI) and Machine Learning (ML) has become increasingly important in stock market prediction. This project, titled “AI Based Stock Market Analysis and Prediction,” aims to develop an intelligent system that uses machine learning and natural language processing techniques to analyze stock market data and predict future price trends. The system provides users with accurate predictions, sentiment insights, and actionable recommendations such as Buy, Sell, or Hold. By combining numerical data with textual sentiment analysis, the system enhances prediction accuracy and supports better decision-making. The core of the system is based on the XGBoost (Extreme Gradient Boosting) regression algorithm, which is known for its high performance and efficiency in handling structured data. Historical stock data, including open, high, low, close, and volume values, is collected using financial APIs such as Yahoo Finance. This data is preprocessed and transformed using feature engineering techniques to extract meaningful patterns. The processed data is then used to train the machine learning model, enabling it to predict future stock prices based on historical trends. In addition to numerical analysis, the system incorporates sentiment analysis using Natural Language Processing (NLP). Financial news and related textual data are analyzed to determine market sentiment, which is classified as positive, negative, or neutral. Since sentiment plays a significant role in influencing stock prices, integrating it with machine learning predictions improves the overall reliability of the system. The final recommendation is generated by combining the predicted price and sentiment score. The system follows a client-server architecture, where the frontend is developed using React.js to provide an interactive user interface, and the backend is implemented using Flask to handle API requests, data processing, and model execution. SQLite is used as the database to store user information and analysis history. The system also includes authentication features such as OTP verification and secure login using JSON Web Tokens (JWT) to ensure data security. The workflow begins when a user enters a stock ticker symbol. The system fetches historical data, processes it, performs prediction and sentiment analysis, and displays the results, including current price, predicted price, sentiment score, and recommendation. The results are stored for future reference, allowing users to track their analysis history.},
      keywords = {Stock Market Prediction; Sentiment Analysis; Machine Learning; Random Forest; Natural Language Processing; Financial Analytics; Time Series Forecasting; Data Preprocessing; Feature Engineering; Social Media Analysis; News Sentiment; Predictive Modeling; Classification Algorithms; Regression Models; Ensemble Learning; Data Mining; Market Trends; Investor Behavior; Accuracy Improvement; Decision Support Systems},
      month = {October},
  }