Regional Prioritization for Free Nutritious Food Programs through Social Data Integration and Public Sentiment Analysis Using K-Means and NLP

Abstract

This study evaluates Indonesia's Free Nutritious Food Program (MBG) through an innovative dual-method approach combining geospatial clustering and sentiment analysis. Cluster analysis of 38 provinces identified three distinct priority zones: high-priority (Eastern Indonesia), medium-priority (Central Indonesia), and low-priority (Java-Bali-West Sumatra), revealing significant regional disparities. Parallel sentiment analysis of 1,358 social media posts showed 76.6% negative perceptions dominated by food safety concerns ("poisoning," "toxic"), contrasting with 23.4% positive feedback highlighting nutritional benefits. The study makes three key contributions: First, it demonstrates the disconnect between regional needs and implementation quality. Second, it introduces an integrated monitoring framework combining cluster mapping with real-time sentiment tracking. Third, it proposes actionable solutions including a rapid-response task force and targeted communication strategies. These findings provide policymakers with evidence-based tools to simultaneously address geographical inequities and improve program execution in nutrition interventions.