Enhancing Financial Market Analysis and Prediction with Emotion Corpora and News Co-Occurrence Network

This study employs an improved natural language processing algorithm to analyze over 500,000 financial news articles from sixteen major sources across 12 sectors, with the top 10 companies in each sector. The analysis identifies shifting economic activity based on emotional sentiment and develops a co-occurrence network show relationships between even sectors. created corpus identify emotions news. method identified 18 additional beyond what was previously analyzed. researchers labeled terms Investopedia validate categorization performance of new method. Using algorithm, we analyzed how relate market movement pairs companies. We found moderate correlation (above 60%) emotion movement. To this finding, further checked coefficients alone, that consumer discretionary, staples, financials, industrials, technology sectors showed similar trends. Our findings suggest provide valuable insights for prediction. technical framework developed can be integrated into larger investment strategy, enabling organizations potential opportunities develop informed strategies. derived model may leveraged by strengthen their risk management functions, making it asset within comprehensive strategy.