Title : A Time-Aware, Deep Learning-Powered Food Recommendation System
Author : P.Goutham Kumar, Deverakonda Mallikerjuna, Kumbha Ramu, Shaik Abeed Basha
Abstract :
It is widely accepted that food recommender-systems can be useful in encouraging people to adopt better dietary behaviours. This paper's goal is to create a new mixed food recommender-system that improves upon the flaws of existing ones by taking into account factors like time, user and food preferences, and community context. The suggested procedure has two parts: the first is a user-based suggestion, and the second is a recommendation based on the food's substance. In the first stage, users and food items are clustered using graph clustering, and in the second stage, users and food items are clustered using a deep-learning based method. In addition, a comprehensive method is used to enhance the suggestion quality by taking into consideration time and user-community associated problems. We used five different performance metrics—Accuracy, Precision, Recall, F1, and NDCG—to evaluate our model against a collection of state-of-the-art recommender-syste
