Evolutionary algorithm optimizes the intervention mechanism of sports content recommendation system on health behavior

The sports content recommendation system based on evolutionary algorithm effectively promotes the improvement of health behavior through personalized recommendation. This study designed a system framework including data collection, user portrait construction, recommendation algorithm and intervention effect evaluation, and used genetic algorithm to optimize the recommended content, taking into account user participation, health goal achievement rate and content diversity. The experimental results show that compared with the deep learning recommendation system, the system improves the recommendation accuracy by about 6.5%, significantly improves the user's sports participation and health indicators, and the average daily exercise time of the experimental group increases from 33.1 minutes to 48.7 minutes, and the heart rate compliance rate increases from 0.65 to 0.79. Through the quantitative evaluation of intervention intensity and behavior persistence indicators, the superiority of the system in health behavior intervention is verified, providing an innovative path for intelligent health management.