Healthy Food Recommendation Using a Time-Aware Community Detection Approach and Reliability Measurement

Abstract Food recommendation systems have been increasingly developed in online food services to make recommendations users according their previous diets. Although unhealthy diets may cause challenging diseases such as diabetes, cancer, and premature heart diseases, most of the neglect considering health factors process. This emphasizes importance reliability from content perspective. paper proposes a new system based on health-aware measurement. In particular, we develop time-aware community detection approach that groups into disjoint sets utilizes identified communities nearest neighbors set rating prediction. Then, novel measurement is introduced by both accuracy criteria predictions evaluate predicted ratings. Also, unreliable are recalculated removing ineffective set. Finally, utilized generate list foods recommendations. Different experiments crawled dataset demonstrate proposed method enhances performance around 7.63%, 6.97%, 7.37%, 15.09%, 16.17% precision, recall, F1, normalized discounted cumulative gain (NDCG), metrics, respectively, compared second-best model.