School physical education can only go as far as the quality of PE teachers goes. An all-inclusive set of assessment indices for the quality of school physical education instruction is necessary since so many variables influence this metric. The assessment index method for physical education teaching quality is becoming more complex as the evaluation period for physical education teaching quality changes. Extracting actionable insights from mountainous assessment datasets is crucial. In this study, we examine the evaluation index system for contemporary physical education teaching quality based on higher physical education, with a computer cloud network as the operational platform and data mining technology's decision tree used. Our university's two undergraduate computer science (class B) and business administration (class A) courses are used as the subjects of this study. At the conclusion of the first semester, students take an exam using the % system as their starting point. Two hundred male and female students, ranging in age from eighteen to twenty-five, were surveyed by questionnaire at the conclusion of the semester, and the second exam was administered as the state following the transfer. With an effective rate of 90.6%, 1935 valid questionnaires were gathered after the survey. Class A averaged 88.63 points in the first semester and Class B 82.73 points, whereas in the second semester, Class A averaged 84.14 points and Class B averaged 81.66 points. After just one semester of instruction, class A's average score dropped 4.49 points, while class B's dropped only 0.07. By using a decision tree algorithm in a thorough review, the impact of instructors' lessons may be more objectively and scientifically assessed. Incorporating data mining decision-making technology into the sports teaching quality evaluation index improves the quality of sports teaching evaluations and promotes theoretical research on sports quality assessment systems.