Integration of machine learning technologies in food flavor research: Current applications, challenges, and future perspectives

Flavor characteristics significantly influence consumer food preferences and purchasing behaviors, constituting a vital research domain in food science. While traditional flavor assessment approaches rely primarily on sensory evaluations and instrumental analyses, they face inherent limitations in processing large-scale datasets and generating comprehensive insights. Advanced machine learning (ML) models have revolutionized flavor research through their high-precision predictive capabilities, effectively addressing conventional methodological constraints. These computational approaches enable sophisticated and efficient flavor analysis by integrating multiple data dimensions, including chemical composition (volatile and non-volatile compounds), sensory attributes (taste, aroma, texture), temporal dynamics (flavor release patterns), and consumer responses. ML models demonstrate remarkable capability in simultaneously processing diverse data types, such as gas chromatography-mass spectrometry results, sensory panel evaluations, and real-time flavor release measurements, to predict consumer preferences and optimize flavor formulations. This review examines state-of-the-art ML applications in flavor science, emphasizing crucial areas such as flavor database development, intelligent sensory detection, and food traceability. Through systematic analysis of contemporary ML algorithms, this study critically evaluates their potential and limitations in decoding complex flavor dynamics, providing valuable insights for both researchers and industry practitioners while identifying promising directions for future technological innovations in food flavor analysis and prediction. The comprehensive synthesis presented here represents a significant contribution to the field by establishing a theoretical framework for ML-driven flavor research and offering practical guidelines for the implementation of computational approaches in food flavor analysis.