Reform of Instructional Assessment System and Cultivation of Students' Core Literacy

This article aims to explore the shortcomings of traditional instructional assessment system, and proposes a multi-dimensional student assessment system based on back propagation neural network (BPNN). By constructing a multi-dimensional assessment index system, this article uses BPNN algorithm to predict and analyze the sample data. Methodologically, abundant student data are collected and a comprehensive assessment index system is constructed. Through BPNN training and prediction, the results are compared with the actual and teachers' subjective scores. The results show that the assessment algorithm based on BPNN has high prediction accuracy and strong consistency with teachers' subjective scores. The reformed assessment system has significantly improved students' core literacy, especially in innovative ability and critical thinking. The multi-dimensional student assessment system based on BPNN is helpful to promote students' comprehensive development and provides a new idea for the reform of instructional assessment system. In the future, we will continue to optimize the algorithm and expand its application to better serve the development of education.