English is a world language, and the ability to use English plays an important role in the improvement of college students' comprehensive quality and career development. However, quite a lot of Chinese college students feel that English learning is difficult; it is difficult to understand the learning materials, and they cannot effectively improve their English ability. This study uses a convolutional neural network to evaluate the readability of English reading materials. It provides students with English reading materials of suitable difficulty based on their English reading ability so as to improve the effect of English learning. Aiming at the high dispersion of students' English reading level, a text readability evaluation model for English reading textbooks based on deep learning is designed. First, the legibility dataset is constructed based on college English textbooks; second, the TextCNN text legibility evaluation model is constructed; finally, the model training is completed through parameter adjustment and optimization, and the evaluation accuracy rate on the self-built dataset reaches 90%. We use the text readability method based on TextCNN model to conduct experimental teaching, and divided the two groups into comparative experiments. The experimental results showed that the reading level and reading interest of students in the experimental group were significantly improved, which proved that the text readability evaluation method based on deep learning was scientific and effective. In addition, we will further expand the capacity of the English legibility dataset and invite more university classes and students to participate in comparative experiments to improve the generality of the model.
Keywords: Deep learning; English reading; Readability assessment; TextCNN.
© 2024 Qin and Irshad.