Data Governance Analysis for Predicting Fresh Fruit Bunch (FFB) Accuracy Using COBIT 2019 and the Random Forest Algorithm

Abstract

Fresh Fruit Bunch (FFB) production is a critical factor in the palm oil industry, as accurate prediction of harvest outcomes directly influences operational planning, resource allocation, and supply chain efficiency. However, the quality and governance of agricultural data often present challenges that can reduce the reliability of predictive models. This study aims to analyze data governance practices using the COBIT 2019 framework and evaluate their contribution to improving the accuracy of Fresh Fruit Bunch (FFB) predictions through the Random Forest algorithm. The research adopts a quantitative approach by assessing selected COBIT 2019 governance and management objectives related to data quality, information management, and decision-making processes. Data were collected from plantation operational records, production reports, and stakeholder questionnaires. The governance assessment was conducted to determine the capability level of existing data management practices, while Random Forest was employed to develop a predictive model for FFB production accuracy. The findings indicate that effective data governance significantly contributes to the quality, consistency, and reliability of plantation data used for machine learning processes. The capability assessment shows that the evaluated COBIT 2019 domains achieved an average capability level of 3.4, indicating that data governance processes are well established and consistently implemented. Furthermore, the Random Forest model demonstrated strong predictive performance, achieving an accuracy rate of 94.2%, precision of 93.6%, recall of 92.8%, and F1-score of 93.2%. The integration of COBIT 2019-based data governance and machine learning techniques provides a structured approach to enhancing predictive accuracy and supporting data-driven decision-making in the palm oil sector. This study highlights the importance of aligning governance frameworks with advanced analytics to improve operational effectiveness and sustainable plantation management.