Sports Training Recommendation Method under the Background of Data Analysis

With the rapid development of the network information age, the degree of information redundancy is increasing. In today's society, various resources are not only rich in categories, but also very complicated in content, so that it is difficult for athletes to efficiently and accurately screen out training items that meet their requirements in a short time. The emergence of recommendation algorithms has effectively alleviated this problem to a large extent. Recommendation algorithms have been applied to more and more fields. Collaborative filtering algorithms stand out among many algorithms due to their conciseness, ease of implementation and good recommendation effect. They have won unanimous approval from experts, scholars and developers, and then began to be applied on a large scale. People's study and life. In view of the above-mentioned problems, this paper studies the traditional collaborative filtering recommendation algorithm and proposes tag-based collaborative filtering recommendation algorithm that incorporates time factors, and a tag-based collaborative filtering recommendation that integrates athletes' interest. Athlete's preference is obtained by constructing an athlete scoring matrix, and then the label weight calculated by the TF-IDF method is combined to construct an athlete's interest distribution model. On this basis, the time factor and the interest of athletes are integrated to construct an athlete similarity model. Through experiments, it can be seen that the newly proposed collaborative filtering algorithm that integrates the time factor.