With the increase in soybean trade between countries, the intentional mislabeling of the origin of soybeans has become a serious problem worldwide. In this study, metabolic profiling of soybeans from the Republic of Korea and China was performed by nuclear magnetic resonance (NMR) spectroscopy coupled with multivariate statistical analysis to predict the geographical origin of soybeans. The optimal orthogonal partial least squares-discriminant analysis (OPLS-DA) model was obtained using total area normalization and unit variance (UV) scaling, without applying the variable influences on projection (VIP) cut-off value, resulting in 96.9% sensitivity, 94.4% specificity, and 95.6% accuracy in the leave-one-out cross validation (LOO-CV) test for discriminating between Korean and Chinese soybeans. Soybeans from the northeastern, middle, and southern regions of China were successfully differentiated by standardized area normalization and UV scaling with a VIP cut-off value of 1.0, resulting in 100% sensitivity, 91.7%–100% specificity, and 94.4%–100% accuracy in a LOO-CV test. The methods employed in this study can be used to obtain essential information for the authentication of soybean samples from diverse geographical locations in future studies.

Rice (Oryza sativa L.) is a widely consumed food source, and its geographical origin has long been a subject of discussion. In our study, we collected 44 and 20 rice samples from different regions of the Republic of Korea and China, respectively, of which 35 and 29 samples were of white and brown rice, respectively. These samples were analyzed using nuclear magnetic resonance (NMR) spectroscopy, followed by analyses with various data normalization and scaling methods. Then, leave-one-out cross-validation (LOOCV) and external validation were employed to evaluate various machine learning algorithms. Total area normalization, with unit variance and Pareto scaling for white and brown rice samples, respectively, was determined as the best pre-processing method in orthogonal partial least squares–discriminant analysis. Among the various tested algorithms, support vector machine (SVM) was the best algorithm for predicting the geographical origin of white and brown rice, with an accuracy of 0.99 and 0.96, respectively. In external validation, the SVM-based prediction model for white and brown rice showed good performance, with an accuracy of 1.0. The results of this study suggest the potential application of machine learning techniques based on NMR data for the differentiation and prediction of diverse geographical origins of white and brown rice.