<![CDATA[Effectiveness of Indonesia’s Free Nutritious Meal Program (MBG) is influenced not only by operational efficiency but also by public acceptance and engagement. This study proposes a data-driven marketing framework integrating Artificial Intelligence (AI) and consumer behavior analysis to enhance program effectiveness. A quantitative and computational approach is employed using secondary and simulated data (N = 1,250), incorporating behavioral and service-related variables such as awareness, trust, perceived benefit, and accessibility. Machine learning techniques, including K-Means clustering for segmentation and Random Forest and XGBoost for predictive modeling, are applied to analyze and predict program acceptance. The results show that the Random Forest model achieves an accuracy of 89.3%, precision of 87.6%, recall of 88.9%, and F1-score of 88.2%, outperforming baseline models. Feature importance analysis indicates that awareness (0.247), trust (0.198), and accessibility (0.158) are the most influential factors, contributing nearly 45% of the model’s predictive power. Segmentation analysis identifies three consumer groups: high acceptance (34.7%), medium acceptance (38.5%), and low acceptance (26.8%), with the medium segment representing the most strategic target for intervention. Furthermore, sentiment analysis in 2025 reveals a dominant positive perception (60.8%), followed by neutral (24.3%) and negative (14.9%) responses, with a gradual increase in positive sentiment over time. The integration of predictive modeling, segmentation, and sentiment analysis enables targeted marketing strategies that improve engagement by up to 18.6%. This study contributes to bridging marketing management and computer science by providing an explainable and adaptive AI-driven framework for optimizing large-scale public programs]]>
