This paper aims to study the evaluation methods and strategies of physical education (PE) classroom teaching effect based on big data. This paper introduces the main elements of PE teaching, and adopts diversified allocation modes of PE teachers to reform the PE teaching mode. Moreover, starting from promoting lifelong PE awareness, this paper chooses reasonable PE teaching contents to deepen students’ ideological understanding and cultivate students’ comprehensive PE quality. In addition, under the optimized teaching mode, this paper introduces test analysis methods to analyze the learning evaluation of PE teaching effects, and constructs the multi-learning evaluation model function of diversified PE teaching effect to optimize the design of the multi-learning evaluation model of diversified PE teaching effect. Finally, this paper verifies the feasibility and application advantages of using artificial intelligence technology to evaluate the effect of PE.