This project introduces a real-time market sentiment analysis system that aims to forecast stock price movements by examining news articles through machine learning methods. It is based on the idea that stock prices are shaped not only by financial indicators but also by public opinion reflected in news and media content. The system processes and categorizes news articles into sentiment classes using models such as Naive Bayes and Random Forest, then generates a sentiment score which is compared with actual stock market trends to uncover meaningful relationships. By combining sentiment insights with market data, the approach enhances prediction accuracy and assists investors in making smarter, data-driven decisions. It also enables continuous monitoring of market sentiment, allowing quicker reactions to changes in market conditions. Additionally, the system highlights how machine learning can be effectively applied in financial prediction and decision support tools while offering a scalable framework capable of handling large volumes of news data. Overall, it provides valuable insights for investors, analysts, and automated trading systems, emphasizing that integrating sentiment analysis with conventional financial techniques can lead to more reliable predictions and improved investment strategies.