The dynamics of the Indonesian capital market are increasingly influenced by information flow in the digital era. PT Bank Central Asia Tbk (BBCA), as a key market proxy, experienced price volatility in 2024–2025 despite solid fundamentals, indicating the influence of market psychology. This study aims to analyze the effect of stock market news sentiment on BBCA stock prices and test the effectiveness of the Multinomial Naive Bayes algorithm. Using a text mining approach, 5,000 economic news articles (2020–2025) were processed using TF-IDF and classified into positive, negative, and neutral sentiments. The results show the model achieved 92.4% accuracy with 89% precision for negative sentiment detection. Pearson correlation analysis revealed a strong positive relationship (r = 0.78) between daily sentiment scores and the following day's closing prices. The study concludes that news sentiment is a valid leading indicator for stock movements. The Naive Bayes algorithm proved efficient for financial text analysis, offering a viable tool for investor risk mitigation.
Copyrights © 2026