Characterization and Discrimination of Volatile Compounds in Honeysuckle From Different Plots Using E‐Tongue, E‐Nose, GC‐IMS, and HS‐SPME‐GC–MS Combined With Random Forest Algorithm

Honeysuckle samples from various plots in Pingyi County, Linyi City were selected and analyzed by integrating electronic tongue (E-tongue), electronic nose (E-nose), HS-SPME-GC-MS, and GC-IMS technologies with multivariate statistical analysis methods, a systematic analysis of the flavor characteristics and plot differentiation patterns of volatile organic compounds (VOCs) was conducted in the study. The results indicated that the E-tongue provided distinct sensor response patterns that correlated with taste attributes such as sweetness and bitterness, effectively discriminating samples from different plots. E-nose analysis further showed that aroma-related response intensities exhibited significant plot-based variation. Through HS-SPME-GC-MS and GC-IMS techniques, a total of 203 volatile components were identified. Based on HS-SPME-GC-MS data, using VIP > 1 and OAV ≥ 1 as screening criteria, seven key volatile markers were determined. From the GC-IMS data, 24 characteristic compounds were screened out using the criterion of VIP > 1, with aldehydes comprising the highest proportion. Furthermore, the random forest (RF) algorithm was employed to integrate these two sets of data. Monte Carlo cross-validation showed that the model accuracy reached 1.0, successfully identifying 15 key volatile compounds. This study provided a theoretical foundation and data support for optimizing the quality evaluation system of honeysuckle and innovating origin traceability technology through multi-dimensional flavor analysis and key marker screening.