Datasets S1-S7 for "The interaction law of odor intensity, pleasantness, and quality of binary mixtures in the process of olfactory perception".

The study of the olfactory perception relationship among odor intensity, pleasantness, and quality (IPQ) of binary mixtures, is always a hot topic among flavor chemists, perfumers, psychophysicists, et al. On the base of extensive sensory evaluation data, the relationships among IPQ within binary mixtures were investigated. First, models about odor intensity and pleasantness were analyzed. The goodness of fit (R²), root mean square error (RMSE), generalization ability (represented by the index of agreement, IA), and predictive ability (represented by Theil U statistic 1, U1) were calculated, with R² = 0.7229 ~ 0.7899, RMSE = 0.5116 ~ 0.5700, IA = 0.8354 ~ 0.8829 and U1 = 0.0026 ~ 0.0028. Although heterogeneous perception models were established to correct the errors of existing models, the error had not been corrected in the pleasantness partially additive situation. Then, to address this issue, the relationship between odor pleasantness and quality was studied. A regression model by correlating odor quality descriptors with the pleasantness of binary mixtures was established by using an orthogonal partial least squares analysis. This model facilitated the identification of a correction factor for variations in the pleasantness of binary mixtures. Using this correction factor, a binary mixture's pleasantness change caused by the change in the quality of the binary mixture can be predicted. Finally, the universal olfactory perception law of aroma intensity, pleasantness, and quality of binary mixtures were obtained, with R² = 0.7904 ~ 0.8179, RMSE = 0.4040 ~ 0.4158, IA = 0.9375 ~ 0.9399 and U1 = 0.002.