Curriculum System Design and Teaching Mode Innovation of English for Specialized Purposes Platform

Abstract

The current status of ESP teaching research is not optimistic, failing to eliminate the limitations and constraints of curriculum design and teaching mode. This study first utilizes the forward and back propagation algorithms in the deep learning network to extract features from the student sample data. The clustering center of the sample data and the weight value of each feature are calculated for the clustering results using the phase dissimilarity calculation of the improved K-prototypes algorithm. Through the analysis of the students’ motivation to learn English in a specific school, it is found that the factor of “achievement motivation” accounts for the most significant proportion, which is 55.80%, and most of the students’ specific motivation to learn English for particular purposes is “to obtain the graduation certificate”, which accounts for 95.3% of the students’ motivation. Most students were motivated to learn English for particular purposes mainly by obtaining a diploma, accounting for 95.3%. Among them, students in category 2 have a higher value of learning characteristics, with an average of 0.77. In contrast, students in category 5 have the lowest value of knowledge mastery, with an average of 0.26. This study lays a data foundation and support for the design and innovation of the English for Specialized Purposes platform. At the same time, the innovative teaching mode of English for Specialized Purposes proposed in this paper based on the CDIO teaching model provides a reference for teaching English for Specialized Purposes in the future.

Keywords

• k-prototypes algorithm
• deep learning network
• feature extraction
• cluster analysis
• specialized english
• 03b22