Market2Dish: Health-aware Food Recommendation

With the rising incidence of some diseases, such as obesity and diabetes, healthy diet is arousing increasing attention. However, most existing food-related research efforts focus on recipe retrieval, user-preference-based food recommendation, cooking assistance, or nutrition calorie estimation dishes, ignoring personalized health-aware recommendation. Therefore, in this work, we present a recommendation scheme, namely, Market2Dish, mapping ingredients displayed market to dishes eaten at home. The proposed scheme comprises three components, user health profiling, In particular, retrieval aims acquire available users then retrieve candidates from large-scale dataset. User profiling characterize conditions by capturing textual health-related information crawled social networks. Specifically, solve issue that extremely sparse, incorporate word-class interaction mechanism into deep model learn fine-grained correlations between tweets pre-defined concepts. For novel category-aware hierarchical memory network–based recommender user-recipe interactions for better Moreover, extensive experiments demonstrate effectiveness scheme.