Background: Electoral participation reflects the quality of democracy, particularly in rural communities with diverse socioeconomic structures. Specific Background: In Sidoarjo Regency, disparities in participation levels among villages suggest that local development factors play a crucial role. Knowledge Gap: Previous models only used demographic attributes without integrating the multidimensional Village Development Index (IDM) indicators. Aims: This study aims to construct a predictive model of voter participation using the Naïve Bayes classification algorithm based on IDM data. Results: By applying preprocessing, feature selection, and probabilistic classification to 48 attributes of IDM, the model achieved 78.65% accuracy, 79% precision, 76% recall, and 77% F1-score, revealing that education, health, and accessibility variables are key predictors. Novelty: Unlike prior research, this work combines social, economic, and ecological IDM dimensions with an open-source Python-based approach for transparent model validation. Implications: The findings demonstrate the feasibility of data-driven governance tools for mapping electoral participation and can support strategic planning to improve civic engagement in rural Indonesia.Highlights:• Uses IDM indicators to predict election participation• Naïve Bayes model achieves 78.65% accuracy• Supports data-driven democratic planning