A Sentiment-Augmented Machine Learning Approach to Forecasting IHSG Prices Using XGBoost

Stocks are a commonly used investment instrument, representing ownership in a company, and offering opportunities for investors to gain profits through the appreciation of stock value as well as dividend distribution. As one of the main financial assets, stocks are also influenced by various external factors, such as economic conditions, government policies, and market sentiment. All of these factors play a crucial role in determining stock price movements. This study integrates sentiment analysis with the XGBoost algorithm to predict IHSG stock prices. By utilizing historical stock data and sentiment derived from financial news, the study evaluates the impact of sentiment data integration on prediction accuracy. Three types of returns (absolute, relative, and logarithmic) and five sentiment scenarios were employed to assess the contribution of sentiment features to the prediction model. The results indicate that sentiment integration consistently improves the predictive performance of the model compared to using historical data alone. Among the tested scenarios, Scenario 2, 4, and 5 demonstrated the best performance, with an RMSE value of 0.009163 and an MAE value of 0.007432, using the logarithmic return type. These findings suggest that incorporating sentiment features into predictive models can enhance the accuracy of stock price predictions and highlight the potential of Natural Language Processing (NLP) and Large Language Models (LLMs) in stock market analysis.