다양한 문화가 결합된 음식

Abstract Navigating cross-cultural food choices is complex, influenced by cultural nuances and various factors, with flavor playing a crucial role. Understanding cultural flavor preferences helps individuals make informed food choices in cross-cultural contexts. We examined flavor differences across China, the US, and Germany, as well as consistent flavor preference patterns using online recipes from prominent recipe portals. Distinct from applying traditional food pairing theory, we directly mapped ingredients to their individual flavor compounds using an authorized database. This allowed us to analyze cultural flavor preferences at the molecular level and conduct machine learning experiments on 25,000 recipes from each culture to reveal flavor-based distinctions. The classifier, trained on these flavor compounds, achieved 77% accuracy in discriminating recipes by country in a three-class classification task, where random choice would yield 33.3% accuracy. Additionally, using user interaction data on appreciation metrics from each recipe portal (e.g., recipe ratings), we selected the top 10% and bottom 10% of recipes as proxies for appreciated and less appreciated recipes, respectively. Models trained within each portal discriminated between the two groups, reaching a maximum accuracy of 66%, while random selection would result in a baseline accuracy of 50%. We also explored cross-cultural preferences by applying classifiers trained on one culture to recipes from other cultures. While the cross-cultural performance was modest (specifically, a max accuracy of 54% was obtained when predicting food preferences ofthe USusers with models trained on the Chinesedata), the results indicate potential shared flavor patterns, especially between Chinese and US recipes, which show similarities, while German preferences differ. Exploratory analyses further validated these findings: we constructed ingredient networks based on co-occurrence relationships to label recipes as savory or sweet, and clustered the flavor profiles of compounds as sweet or non-sweet. These analyses showed opposing trends in sweet vs. non-sweet/savory appreciation between US and German users, supporting the machine learning results. Although our findings are likely to be influenced by biases in online data sources and the limitations of data-driven methods, they may still highlight meaningful cultural differences and shared flavor preferences. These insights offer potential for developing food recommender systems that cater to cross-cultural contexts. Keywords: food preferences, flavor compounds, food cultures, food recommender systems 1. Introduction Deciding what to eat is a complex process shaped by many contextual factors [1,2,3,4], among which culture plays a significant role [5,6,7,8]. Varied cultural and regional groups have different eating habits [9,10,11,12,13]. Individuals commonly find it is easier to make food choices within their familiar cultural contexts due to a sense of identity and familiarity [14,15,16], while in cross-cultural scenarios such as travel or migration, they often navigate between exploring new cuisines and maintaining ties to their own food culture [17,18], adding further complexity to food decision-making. Food recommender systems have emerged to help users manage the overwhelming number of food options by tailoring suggestions to personal preferences. Previous work has also emphasized the importance of incorporating cultural elements into recommendation systems [19]. Nevertheless, understanding cross-cultural flavor preferences remains an ongoing challenge. Traditional anthropological methods, such as case studies and questionnaires, have provided valuable insights into global dietary patterns. By focusing on small groups within specific ethnic communities, anthropologists have not only uncovered the unique characteristics of each culture [20,21,22], but also explored how various factors—such as environmental contexts [14,23,24,25,26], rituals [1,27,28], religions [29,30,31], and aesthetic ideals [32,33,34,35,36]—shape food practices, either individually or in combination. However, in-depth qualitative approaches are expensive and time-consuming. Over the past decade, online recipe portals have increasingly become a testbed for studying cross-cultural dietary patterns. By applying computational methods on large-scale online sources, such as online recipes, and users’ activity on social media, researchers can identify broad differences in food preferences with quantitative evidence. For example, research leveraging large-scale recipe datasets demonstrated significant variations in food choices across cultures. Kim and Chung [12] found that recipes from 22 countries feature distinct ingredients, such as ginger and soy sauce in Asian recipes and olive oil in European cuisines. Similarly, Sajadmanesh et al. [37] found that user-generated recipes from 200 countries have unique ingredient patterns, allowing recipes fr