COMPARATIVE ANALYSIS OF SENTIMENT ANALYSIS MODELS ON BANKING INVESTMENT IMPACT BY MACHINE LEARNING ALGORITHM

This study investigates the application of sentiment analysis on financial news to predict its impact on banking investments, comparing the performance of various machine learning and deep learning models. Given the complexities and volume of financial news, leveraging sentiment analysis provides valuable insights into investor sentiment, which can influence market dynamics and banking sector investments. We conducted a comparative analysis of six sentiment analysis models: Naïve Bayes, Support Vector Machine (SVM), Random Forest, Gradient Boosting, Long Short-Term Memory (LSTM), and BERT fine-tuning. Our results reveal that while traditional models like Naïve Bayes and SVM provide a foundational accuracy level, they lag behind more advanced methods in both sentiment accuracy and correlation with investment trends. Notably, the BERT fine-tuning model demonstrated superior performance, achieving the highest accuracy at 89.4% and the strongest correlation with banking investment outcomes (Pearson’s R = 0.68). These findings highlight the effectiveness of deep learning models, especially transformer-based models like BERT, in handling the linguistic nuances and predictive challenges inherent in financial sentiment analysis. By offering insights into the relationship between financial news sentiment and banking investments, this study underscores the potential of sentiment analysis as a tool for informed decision-making and enhanced forecasting in the financial sector.