A dog food recommendation system based on nutrient suitability

The demand for a food recommendation service for dogs has rapidly increased with the increasing number of pet owners, because it is generally difficult for dog owners to find food that is perfectly suitable for their dogs' health condition. The purpose of this study is to develop an algorithm for recommending dog food that contains appropriate nutrients based on the physical and health conditions of the dogs. This study proposes a nutrient profiling-based recommendation algorithm (NRA) for dog food. The proposed algorithm tries to recommend appropriate or inappropriate dog food by using collective intelligence based on user experience and the prior knowledge of experts. Based on the physical and health status of dogs, this study extracts which nutrients are necessary for the dogs and recommends the most suitable dog food containing these nutrients. A performance evaluation was implemented in terms of recall, precision, F1 and AUC. As a result of the performance evaluation, the AUC performance of this NRA is 20% higher than k-NN and 9.7% higher than the SVD model. In addition, the NRA proved to be an evolving system in which the performance of recommendations improves as users' feedback accumulates.