Exploration and Practice of College English Reading Teaching Reform in the Context of Deep Learning

Abstract

Abstract To improve the quality and effectiveness of learners’ English reading learning, this paper proposes a fuzzy hierarchical analysis method incorporating deviance entropy (DE-FAHP), which improves the differentiation quality of the fuzzy judgment matrix and enhances the accuracy of weight calculation of each influencing factor by introducing deviance entropy model. Finally, the comprehensive difficulty quantification model of English reading comprehension objective questions is determined to provide an efficient and accurate method for calculating the parameters of the subsequent English reading comprehension standardized question bank. Finally, this paper proposes a hybrid testing method that combines CAT’s online grouping technology with MST’s staged presentation of test questions to construct the English reading comprehension adaptive testing system ca-MST. Through system validity testing, students in the experimental class improved more significantly, with only question 20 in the experimental class scoring below 4.0, while most of the scores in the control class were below 4.0. The mean reading score of the experimental class was 76.31, with a standard deviation of 11.16, while the mean reading score of the control class was 69, with a standard deviation of 18.76. The comparison showed that the reading scores of the experimental class taught with the reading comprehension adaptive testing system were higher than those of the control class taught with the traditional model, and the differences within the class were smaller.