摘要：Lifestyle and personality can often be discovered through daily food choices. For instance, influence through social groups, food documentaries or the desire for fitness/weight loss can often create a shift in people’s food choices. These changes in diet can highlight important life events, an overall lifestyle change or an individual’s culture. One way to study people’s cultural food patterns is to study the language they use to share recipes. This is especially important as new recipes can now be discovered with the click of a button as opposed to traditional recipe books, creating online recipe sharing communities. Users can share their most personal or newly learned recipes as well as create a dialogue of feedback or suggestions. Given that food is a major component of our physical and emotional well-being, the focus of this research was to investigate food communication among online communities. In study 1, topic modeling analysis was performed on recipes taken from the popular social networking sites allrecipes.com (N = 32,944) to identify 11 major themes around the consumption of food. Next, study 2, a replication study, used identical topic modeling analysis on the SNS food.com (N = 190,808) to identify 13 themes. Implications of these groupings, as well as the social and personal settings users reported trying the recipes, are discussed, along with limitations and suggestions for further study.
关键词：meaning extraction helper;recipe;online community;text analysis;big data
