A Study on Financial Market Sentiment Analysis and Investment Strategy Formulation Based on Bayesian Networks

Abstract

This study investigates the application of Bayesian networks for financial market sentiment analysis and investment strategy formulation. By integrating sentiment analysis techniques with probabilistic modeling frameworks, we develop a Bayesian network model that captures the complex interplay between sentiment indicators, economic fundamentals, and market variables. Historical market data spanning a five-year period is utilized to train and validate the model, with sentiment scores derived from news articles and social media feeds serving as key inputs. Performance evaluation metrics, including accuracy, precision, recall, F1-score, and area under the ROC curve, are computed to assess the model's predictive capability and discriminative power. The experimental results demonstrate the effectiveness of the Bayesian network model in accurately capturing sentiment trends and predicting market outcomes. Furthermore, the formulated investment strategies based on the Bayesian network analysis outperform traditional benchmarks in terms of annualized return, Sharpe ratio, and maximum drawdown. These findings highlight the practical utility of sentiment-driven investment approaches and underscore the importance of leveraging advanced quantitative techniques to enhance investment decision-making processes in today's dynamic financial markets.