Mapping machine learning applications across the food value chain through a systematic literature review and bibliometric analysis

Food industry plays major role in public health, economic and environmental sustainability. Integration of machine learning (ML) is increasingly reshaping elements like safety protocols, quality management and supply chain optimization, though outcomes vary substantially by context and scale. This paper comprehensively synthesizes the ML applications by employing a systematic literature review (SLR) and a bibliometric analysis of 234 peer-reviewed articles published since 2014 and identifies prominent ML techniques like neural networks, support vector machines, and ensemble techniques. Various applications of these methods are food processing, fraud detection, personalized nutrition, and predictive maintenance. Collaborative pattern along with possible research directions are predicted using keywords co-occurrence and co-authorship network. Explainable AI (XAI), federated learning (FL) along with ethical rules are essential to ensure implementation and navigate AI driven revolution in food sector. Deep learning models have achieved very high classification accuracy in controlled food-quality trials, while early ML-driven dynamic-pricing pilots in SME grocery settings have reported substantial reductions in food waste and modest revenue gains; however, these results are context-specific and remain unconfirmed across broader, resource-constrained SME contexts.