Abstract
Beer consumption behaviors within China exhibited significant regional heterogeneity. To elucidate the specific differences in beer consumer behaviors across different regions and their influencing factors, this study systematically analyzed the sensory preference characteristics of consumers in the Chinese beer market based on machine learning methods, and further revealed the core driving mechanisms influencing their consumption behaviors. By integrating consumer data from different regions, a comprehensive dataset was constructed encompassing sensory attribute evaluations (bitterness, malt flavor, hop aroma, smoothness of mouthfeel, foam characteristics, etc.) and other dimensional consumption behavior variables (brand, beer packaging, etc.). Utilizing an ensemble learning framework (LightGBM), Support Vector Machine (SVM), and decision tree models for feature mining, the study identified important factors influencing the consumption behaviors of Chinese beer consumers. Specifically, consumers in mature and upgrading markets placed greater emphasis on the overall drinking experience and drinkability when purchasing beer, whereas consumers in scale-dominant and mainstream competitive markets considered foam persistence, fineness, and light brown color as core quality indicators. Conversely, consumers in potential growth and emerging cultivation markets demonstrated strong brand orientation. This indicated that the factors influencing beer consumption behaviors varied significantly across regions. Through a data-driven paradigm, this study revealed the underlying regional mechanisms behind consumption decisions in different regional beer markets in China, providing a theoretical foundation and empirical support for cross-regional product customization, precision marketing, and resource optimization.