Research on Digital Transformation of Vocational Education in the Era of Big Data

Abstract

Abstract In the process of development and transformation of vocational education, it is difficult for traditional methods to carry out a digitalized and accurate description of it. Aiming at the construction of the discipline of vocational education, this paper innovates the teaching method by constructing a knowledge graph and uses TF-IDF as the keyword extraction algorithm for entity extraction. To ensure that entity words are recognized accurately, LSTM neural units are utilized to eliminate problems related to gradient vanishing and explosion. On this basis, knowledge graphs are used to integrate heterogeneous data from multiple sources to apply knowledge graphs to the recommender system to create a personalized teaching model to provide schools with a digital transformation method for vocational education. Two different vocational schools tested the model design after it was completed. The analysis of the data revealed that knowledge points experienced an average decrease in error rate of 10-20%. The error rate of knowledge points 10, 11, and 12 did not change much, but the rate of student guessing decreased to 0. The percentage of users who improved their grades in the experimental class using the system by more than 0% was 5.90% higher than that of the normal class, the percentage of those who improved by more than 5% was 5.28% higher, and the percentage of those who had great improvement was 1.90%.