In order to solve the problems of metadata conversion speed and low information utilization of traditional resource integration methods, a data-driven optimization model of physical education curriculum content is proposed. In this paper, we construct a network sports teaching resources metadata set, combine the key attribute matching idea with the design of network sports teaching resources metadata program, and utilize the constructed resource rapid integration platform to integrate the network sports teaching resources metadata. The experimental results show that the information utilization rate of the network sports teaching resources rapid integration method based on key attribute matching is higher than that of the network sports teaching resources integration methods based on data mining technology, Internet of Things and big data technology, and the information utilization rate of the method increases significantly over time, reaching a maximum of 80.235%, which is highly usable.