Sentiment Analysis and Stock Data Prediction Using Financial News Headlines Approach

Abstract

This research provides a web application for stock prediction that analyze financial news sentiment and predicts market performance using machine learning techniques. Machine learning techniques used are vector based, lexicon based analysis and LSTM. For sentiment analysis and stock prediction. With 86% accuracy, the sentiment analysis algorithm categorizes news headlines as positive or negative. The stock prediction model estimates stock performance with a mean absolute inaccuracy of 3.4 percent. When both models are combined, they forecast stock performance with an accuracy of 83%. The system’s potential for usage in the stock market is demonstrated by the results, which provide vital insights into machine learning algorithms for stock data prediction along with sentiment analysis utilizing headlines from financial news.