The Construction and Application of Rural Teachers’ Educational Research Literacy Indicator System in the Context of Deep Learning

Abstract Primary and secondary school teachers are an important group in the development and reform of rural basic education, and the study of rural teachers’ educational research literacy can help improve the quality of rural education. This paper designs an evaluation model of rural teachers’ educational research literacy based on PSO-BP combination algorithm. The BP neural network and particle swarm algorithm are coupled, and the PSO algorithm is used to replace the gradient descent method in the BP neural network to repeatedly optimize the combination of weights and parameters of the BP neural network model until the fitness of the solution is no longer reduced. On this basis, the BP neural network is then used to further precisely optimize the obtained network parameters until the optimal network parameters are searched, and the precise optimal parameter combinations are obtained. The evaluation system of rural teachers’ educational research literacy was constructed by principal component analysis, and three common factors, scientific research ability, scientific research awareness and scientific research ethics, were extracted, and the weights assigned to 26 indicators were calculated according to the factor loadings. In the performance experiment, with the same population size and number of iterations, this paper’s algorithm for R 2 is 0.9991, which is higher than the three baseline models of 0.9834, 0.9875, and 0.9987. With the optimal population size and number of iterations, the regression coefficient is the closest to 1, and the model output values of the test set samples have the smallest relative error to the true value, and the results of the grading are completely consistent. This study provides a scientific reference for accurately assessing the educational research literacy of rural teachers.