MODEL FOR AGRICULTURAL PRODUCTS QUALITY ASSESSMENT (AGRICQUAL)

ABSTRACT

Agriculture is a vital component of Nigeria’s economy, yet the sector undergoes persistent rejection of its exports due to non-compliance with global food safety standards. This study deals with the limitations of manual, reactive quality monitoring by developing AgricQual, an intelligent decision-support system that predicts export compliance for Nigerian agricultural commodities. Adopting a Design Science Research (DSR) paradigm, a dataset of 1,000 observations encompassing eight critical quality indicators was mapped against Codex Alimentarius and European Union (EU) thresholds. Five supervised machine learning models were trained and evaluated. The Random Forest algorithm emerged as the optimal predictive engine, achieving an overall accuracy of 91.67% and a critical recall of 96.74%, effectively minimizing false approvals. The deployed web-based application features a rule-based explanation module, empowering stakeholders with an interpretable, data-driven tool to ensure compliance, reduce economic losses, and foster inclusive economic growth in global trade.

Keywords: Agricultural Export, Machine Learning, Quality Assessment, Random Forest, Decision Support System.