Research and Implementation of Personalized Recommendation Algorithm for Senior Diet Exercise Based on Collaborative Filtering
Xinting Wang, Yifan Song, W. D. Chen, Haiyan Du, Xiaohui Su, H.T. Wang · 2023
In this paper, machine learning methods are introduced to build a personalized recommendation algorithm based on collaborative filtering for meals and exercises for the elderly to meet their personalized needs for meals and exercises. Firstly, a database was established by combining diet and exercise data and medical knowledge, and then an item-based and user-based recommendation algorithm was studied based on collecting personalized preferences of target users, and finally, different weights were assigned to the three variables ICF, UCF and Initial according to their different importance to obtain a fused personalized diet and exercise recommendation algorithm to achieve Top-n personalized recommendation for users. The results show that when the number of users is 500, the accuracy of the personalized meal and exercise recommendation algorithm is 63.1%, which is better than the item-based collaborative filtering algorithm (59.7%) and the user-based collaborative filtering algorithm (52.5%), and the correctness of the fusion algorithm is higher through verification. In the actual test situation, most users were satisfied with the fusion algorithm (60%), indicating that the personalized dietary exercise recommendation algorithm has high reliability, and the personalized dietary exercise recommendation algorithm has been applied to the recommendation system of dietary exercise for the elderly.