<abstract xmlns="http://www.w3.org/1999/xhtml">

<p>Reading is an essential part of English learning and one of the basic elements of language proficiency. This paper proposes a joint analysis model based on multilayer perceptron, which uses an LSTM network model suitable for text processing. Using the BiLSTM network can obtain all the contexts before and after the current word and avoid the problems of gradient disappearance and gradient explosion that exist in ordinary recurrent neural networks. Two multilayer perceptrons are used, one for predicting lexical annotations and the other for predicting the relationship types of dependency arcs, and the model accuracy is further improved by continuously optimizing the parameters in the model. The English corpus is divided into a training set and a test set, and the prediction rate of the model in this paper reaches 96.02% in the English corpus. To test the effectiveness of this method for college English reading instruction, this model was implemented in the experimental class, while the traditional PWP teaching model was still used in the control class. The mean reading post-test score of the experimental class was 31.54, and the mean post-test score of the control class was 26.35. The mean reading post-test score of the experimental class was 7.7 points higher than the mean score of the pre-test. Therefore, the method in this paper is effective in teaching English reading and can improve college students’ reading performance.</p>
</abstract>

Reading is an essential part of English learning and one of the basic elements of language proficiency. This paper proposes a joint analysis model based on multilayer perceptron, which uses an LSTM network model suitable for text processing. Using the BiLSTM network can obtain all the contexts before and after the current word and avoid the problems of gradient disappearance and gradient explosion that exist in ordinary recurrent neural networks. Two multilayer perceptrons are used, one for predicting lexical annotations and the other for predicting the relationship types of dependency arcs, and the model accuracy is further improved by continuously optimizing the parameters in the model. The English corpus is divided into a training set and a test set, and the prediction rate of the model in this paper reaches 96.02% in the English corpus. To test the effectiveness of this method for college English reading instruction, this model was implemented in the experimental class, while the traditional PWP teaching model was still used in the control class. The mean reading post-test score of the experimental class was 31.54, and the mean post-test score of the control class was 26.35. The mean reading post-test score of the experimental class was 7.7 points higher than the mean score of the pre-test. Therefore, the method in this paper is effective in teaching English reading and can improve college students’ reading performance.

Optimization of Innovative Teaching Model of College English Reading Based on Perceptual Machine Model

Applied Mathematics and Nonlinear Sciences

This work is licensed under the Creative Commons Attribution 4.0 International License.

Reading is an essential part of English learning and one of the basic elements of language proficiency. This paper proposes a joint analysis model based on...

Optimization of Innovative Teaching Model of College English Reading Based on Perceptual Machine Model

Publicado en línea: 02 sept 2023

Páginas: -

Recibido: 02 dic 2022

Aceptado: 13 mar 2022

DOI: https://doi.org/10.2478/amns.2023.2.00278

Palabras clave
Multilayer perceptron model, College English, Dependency arcs, LSTM network model, Lexical annotation

© 2023 Lang Wang, published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Optimization of Innovative Teaching Model of College English Reading Based on Perceptual Machine Model

Publicado en línea: 02 sept 2023

Páginas: -

Recibido: 02 dic 2022

Aceptado: 13 mar 2022

DOI: https://doi.org/10.2478/amns.2023.2.00278

Palabras claveMultilayer perceptron model, College English, Dependency arcs, LSTM network model, Lexical annotation

© 2023 Lang Wang, published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Palabras clave
Multilayer perceptron model, College English, Dependency arcs, LSTM network model, Lexical annotation