Making customized recommendations has been a key feature of many websites, and it's just getting bigger as more people have access to vast volumes of data on the internet. When done correctly, giving recommendations based on the individual's preferences rather than trendy products improves consumer satisfaction and will eventually draw more buyers. One more challenge is to deliver healthy recommendations that represent the needs of consumers and preferences and information on the wellbeing of the consumers. This paper proposes a method that categorizes recipes by cuisine and calories that will help to improve rating predictions. We built the proposed Hybrid Hierarchical Clusters Recommender System (HHCRS) with cuisine and calories information and compared its performance with baseline and other minimal recommender systems. The findings showed that our approach can substantially boost prediction and can help to minimize the sparsity of the rating matrix.
Key words: Recipe Recommender System, Cuisines, Calories, Clustering, Artificial Intelligence
|
