Sentiment Analysis of Financial News Article Using TextCNN Algorithm and Web Scrapping

In the world run by finance, Analysing the sentiments of news article that fuels the market trend is critical. This study proposes a Live sentiment analysis model that leverages TextCNN, that achieves a whooping accuracy of 85% in classifying the financial news into positive, negative or neutral sentiments for the market. With the integration of Optical Character Recognition (OCR) and Web-Scraping the research has processed over 30,000 financial news articles for high-quality sentiment extraction. The time-series corelation analysis demonstrates that the 3.7% average market downturn, that highlights the impact of news sentiment on stock movement. These insights help hedge funds, traders and investors to analyse and make a precise decision to invest in the market and it also helps to improving the risk management and strategic formulation.