Research on Corpus-based Graphic Analysis of Ceramic Science and Technology Texts and Its Application in Multimodal English Teaching

Abstract

Abstract With the continuous development of English education, the use of multimodal teaching methods in English instruction has gradually increased. This paper analyzes the characteristics of multimodal discourse and explores the principles of multimodal selection in English teaching. On this basis, a corpus-based graphic analysis framework for ceramic science and technology is constructed. A graphic matching model for ceramic science and technology text is established by combining text feature extraction and graphic feature extraction. The model is optimized using word vector distance and angular similarity measure algorithms, and an English-Chinese bilingual corpus is constructed based on web crawling. Empirical analysis shows that the metrics in text matching in the FK dataset have improved by 0.10% to 2.70% compared to CAAN. Image matching has been enhanced by 1.28% to 2.19%. The average value for both MS and FKMS datasets is 75.85%. Textbook A contains 582 more images with equal relationships than Textbook B and 155 fewer images with unequal relationships. This demonstrates that Textbook A provides a better way for teachers to guide students’ interpretation in multimodal English teaching. This study enhances English teaching and student skill development and also verifies the effectiveness of the graphic analysis method applied in English teaching.