In the era of digitization, concern for health and quality of life has become a top priority. However, maintaining a balanced nutritional lifestyle remains a challenge for many, especially as daily life becomes increasingly hectic. Inadequate and imbalanced dietary habits can lead to various health issues, such as nutritional imbalances, a weakened immune system, and more. Many people have resorted to overusing dietary supplements as meal replacements, causing unwanted side effects on the body. Particularly, choosing suitable dietary regimes is crucial for individuals suffering from various illnesses. To address this issue and support consumers, especially in Vietnam, in selecting meals that match their tastes and nutritional needs while saving time, we have developed a Vietnamese food recommendation system. In this study, we constructed the Vietnamese food dataset - ViFoodRec and processed the data to create a high-quality dataset consisting of the foods dataset with over 5000 data points and the ratings dataset with approximately 180,000 data points. Furthermore, we applied Collaborative Filtering and Content-based Filtering techniques for recommending meals based on users preferences. In both method, Pearson and Cosine are utilized. However, in the context of Content-Based Filtering, we incorporated four additional similarity measures, namely Jaccard, BM25, TfIdf Recommender, and a composite measure
Chi Thanh Dang, Thuy Hong Thi Dang and Quoc An Tran Huynh
Faculty of Information Science and Engineering, University of Information Technology, Vietnam National University, Ho Chi Minh City, Vietnam.