The sensory characteristics of foods have become an important part of measuring the quality of foods, and the descriptive analysis of sensory characteristics is of great significance to the evaluation of the sensory characteristics of foods. Currently, the commonly used ‘static’ sensory descriptive methods include quantitative descriptive analysis (QDA) primarily based on expert panels, and the check-all-that-apply (CATA) and rate-all-that-apply (RATA) methods that are available for experts or consumers to profile the overall sensory characteristics of a product. However, the sensory characteristics of foods can change during consumption, so ‘static’ description methods are unable to fully and accurately represent the sensory characteristics of food products and their delicate changes. Therefore, dynamic sensory description methods such as temporal check-all-that-apply (TCATA), time intensity (TI), and temporal dominance of sensation (TDS), which more accurately and completely describe the sensory characteristics of food products, have attracted great attention. In this article, several common ‘static’ and dynamic sensory description methods and their applications are introduced in order to provide methodological guidance for research on the sensory characteristics of food products.

Flavor substances affect the sensory properties of food and consumer choices, and flavor substance analysis is crucial for improving food quality and developing new products. However, the vast amount of flavor substance data and inappropriate statistical analysis greatly limit the development of this field. Therefore, it is crucial to use new chemometrics methods, such as artificial intelligence algorithms, correctly and reasonably to obtain effective information in this field. In recent years, chemometrics methods have been widely applied in food research. In addition to dimensionality reduction, classification and regression methods, various neural network methods have also emerged in the field of food research. However, a summary of their reasonable application is lacking. Therefore, this article summarizes the statistical analysis methods available to study food flavor, including principal component analysis, linear discriminant analysis, linear regression methods such as partial least squares regression, and nonlinear methods such as fuzzy logic and artificial neural networks, explains their principles and provides application examples. This article aims to provide effective methods and ideas for further research on chemometrics in the field of food flavor.