Research on the Optimization of English Translation Teaching Mode in Colleges and Universities Driven by Knowledge Mapping

Abstract

Due to various factors, the effect of university English translation teaching cannot reach the expected situation. This paper introduces the attention mechanism based on a knowledge graph, builds a graph neural network translation model based on this model, and constructs an English translation teaching model centered on this model. The performance test of the graph neural network model is carried out, and by comparing with other models, it is found that the perplexity of the graph neural network model is only 1.4 with the increase of the number of steps, and the highest value of the graph neural network model is only 28s and the lowest value is 17s in different English translation scenarios. It shows that the graph neural network model has high operational efficiency. The study tested the teaching model based on a graph neural network model with 80 logistics management students, using test papers and analyzing the data using the SPSS 21.0 statistical tool. It was found that the mean difference between the scores of the two classes after the experiment was 4.5675, which is a large difference, and the P=0.045<0.05 obtained from the T-test, indicating that there is a significant difference between the scores of the experimental class and the control class in the post-test. Except for listening comprehension and reading comprehension, there is no significant difference in the test scores of the experimental class compared to the control class in other subjects. The teaching model that utilizes graph neural network translation can improve students’ English translation performance to a certain extent, as demonstrated.