Search

The advancement of AI technology, especially in language processing systems like ChatGPT, creates opportunities for students to elevate their writing skills. This study aimed to explore the relationship between the usability of ChatGPT and the English essay writing proficiency of the BEED students. The study utilized a descriptive correlational design to investigate the relationship. A sample of 58 participants was selected to write an essay with the assistance of ChatGPT. Researchers developed and employed a rubric to assess the quality and ethical use of ChatGPT. Then, the participants answered the USE questionnaire to measure their perceived usability of ChatGPT. The result revealed that BEED students perceived ChatGPT as beneficial in enhancing their essays and found the system easy to use, and learn, and provided satisfying results. The participants also expressed a strong sense of contentment regarding using ChatGPT. The essay writing proficiency assessment revealed that BEED students were proficient writers. However, a Pearson correlation coefficient of 0.042 showed a statistically significant but very weak relationship observed between ChatGPT usability and English essay writing proficiency levels. It can be concluded that ChatGPT may improve students' essay writing proficiency to a minimal extent. Moreover, this study recommended further research to explore ChatGPT's use in other academic outputs, refine the rubrics to include vocabulary, grammar, and sentence structure criteria, and provide training for effective AI tool management among students. Also, policymakers should establish clear policies and guidelines to ensure the responsible and ethical use of ChatGPT, maintaining academic integrity and enhancing writing proficiency.

As the main form of constructing and disseminating disciplinary knowledge, English journal articles have an important linguistic status and theoretical research value. Among them, metadiscourse, as a typical discourse category presented by academic discourse behaviors, has been attracting attention for its discourse features. In order to explore the variation mechanism of metadiscourse verb patterns more effectively. In this paper, we construct a BiLSTM-verb analysis model in intra- and inter-disciplinary perspectives. The computational analysis shows that the authenticity embodiment of metadiscourse is 89%. The interactivity of metadiscourse is 78%. The meta-discourse needs to deal appropriately with the relationship with previous research, and to highlight its own findings and significance. A longitudinal comparison shows that the first paragraph highlights the conceptual interpretation of subject knowledge and the refinement of propositional relationships, with more emphasis on the abstraction and generality of information presentation. Therefore, its variance fit is up to 23%. The diagrams focus most on explaining the concepts of subject knowledge, indicating the tone and level of reasoning of the discourse, and outlining the purpose and results of the study. Therefore, its variance fit is 19%.

The study of metadiscourse word resources is an important way to reveal academic knowledge constructs and genre features in English papers. This paper selects intra- and interdisciplinary English academic papers as data samples, and the text representation is based on the vector space model. The text similarity is calculated using the word movement distance model to complete the clustering based on the density peak, and then the semantic annotation of metadiscourse verb patterns is performed using the conditional random field. Then, the intra- and interdisciplinary metadiscourse annotations are compared to analyze the variation of verb patterns. The intra-disciplinary variation of metadiscourse verb patterns was mainly found using attitude markers, where the log-likelihood ratios of 48.56, 30.25, and 35.64 were obtained for affirmative, critical, and critical declarative attitudes, respectively. In interdisciplinary studies, basic disciplines preferred to use verb patterns as transitions, while applied disciplines preferred to express the author’s point of view. The study of metadiscourse verb patterns effectively enriches the linguistic resources in academic English teaching and deepens the understanding of the characteristics of disciplinary essay genres.

Grammatical error correction (GEC) is an important element in language learning. In this article, based on deep learning, the application of the Transformer model in GEC was briefly introduced. Then, in order to improve the performance of the model on GEC, it was optimized by a generative adversarial network (GAN). Experiments were conducted on two data sets. It was found that the performance of the GAN-combined Transformer model was significantly improved compared to the Transformer model. The F 0.5 value of the optimized model was 53.87 on CoNIL-2014, which was 2.69 larger than the Transformer model; the generalized language evaluation understanding (GLEU) value of the optimized model was 61.77 on JFLEG, which was 8.81 larger than that of the Transformer model. The optimized model also had a favorable correction performance in an actual English essay. The experimental results verify the reliability of the GAN-combined Transformer model on automatic English GEC, suggesting that the model can be further promoted and applied in practice.

Copyright of Journal of Zhejiang University (Science Edition) is the property of Journal of Zhejiang University (Science Edition) Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

This article investigates the lexical and discourse features of English text and discourse with automated computer technologies. Specifically, this article examines the cohesion of English text and discourse with automated computer tools, Coh-Metrix and TEES. Coh-Metrix is a text analysis computer tool that can analyze English text and discourse on various linguistic and psycholinguistic measures of cohesion, readability, and language. Many researchers in the areas of applied linguistics, English education, and language psychology have now extensively used Coh-Metrix to analyze various English texts and textbooks. Recently, the author of this article has developed a new computer tool, TEES which can be applied to evaluate English texts and essays on various linguistic and psycholinguistic measures such as text readability, text cohesion, sentence structure, vocabulary, and text marker scores. Basically, TEES has been developed to evaluate English texts and essays in terms of a standardized norm. In the TEES system, a huge size of corpus was used to construct the standardized norm. This article introduces Coh-Metrix and TEES, and presents some research findings collected from Coh-Metrix studies. [ABSTRACT FROM AUTHOR]