Abstract
Artificial intelligence has been widely used in higher education. Particularly in language teaching, AI has been extensively utilized for material development and adaptations. This article presents the author’s experience of integrating XIPU AI to adjust language input materials, design paraphrasing and summarizing exercises, and create note-completion and vocabulary expansion exercises. Through an evaluation of the effectiveness of AI generated materials, XIPU AI proves to be an instrumental tool in terms of expediting lesson preparation and diversifying language learning activities. However, it is recommended that educators should maintain their discernment and expertise while integrating AI in language teaching and be cautious with the potential risks associated with an excessive reliance on AI.
Introduction
Artificial intelligence has been widely used in higher education. Particularly in language teaching, AI has been extensively utilized for material development and adaptations (Koraishi, 2023). This article aims to provide insights into my experience with using XIPU AI, detailing the specific applications, methods employed, and an assessment of the effectiveness of the generated materials. Additionally, I will discuss the benefits observed for both myself and my students, as well as the potential drawbacks that educators and learners need to be aware of.
Practical applications of XIPU AI
The first kind of material development and adaptation focuses on proficiency levels (Koraishi, 2023) and styles. I used XIPU AI to assist in adjusting teaching materials to better align with my students' backgrounds and language proficiency. This approach aimed to enhance the language acquisition process and foster an awareness of academic style, encouraging the use of advanced vocabulary and complex sentence structures. Curriculum reading materials and articles from the internet, with the same topic but in informal styles, were prepared in advance. To initiate the adaptation process, I created my prompts to guide the AI. For example, “I want you to help me revise the given material at the C1 level and try to use more academic vocabulary, complex sentence structures, and cohesive devices to connect the sentences.” Subsequently, XIPU AI generated texts based on these instructions. The evaluation phase focused on assessing the quality of the materials produced. Criteria considered included the comprehensibility of the adapted version, the degree of academic style exhibited in the revised text, and the extent to which sentence complexity and cohesion were enhanced. I saved the original and the adapted versions in different columns to find out the words or sentences that were well revised. The extracted examples served a dual purpose of showcasing alternative expressions of the same meaning and raising students’ awareness of paraphrasing and summarizing techniques. When presenting the original and adapted versions to students, a visual aid was incorporated using different colors to highlight linguistic changes, such as lexical or syntactic alterations.This approach aimed to enable students to explicitly observe and comprehend the nuances within the generated texts.
Secondly, XIPU AI was utilized to assist in paraphrasing and summarizing activities. Previous research reported students’ difficulties in utilizing paraphrasing and summarizing skills in academic writing (Regala-Flores and Lopez, 2019). As an EAP lecturer for year two students, I also observed persistent challenges among students in paraphrasing and summarizing source information in their own words, leading to poor paraphrasing and academic integrity penalties in their writing coursework. To help students overcome these difficulties, I used XIPU AI to create paraphrasing exercises that incorporate various paraphrasing techniques, such as using synonyms, changing word forms, changing word order, and modifying sentence structures. Sentences containing sophisticated ideas, topic and subject-specific vocabulary, and complex grammatical structures were carefully extracted from curriculum materials. These sentences were then submitted to XIPU AI, which generated paraphrased versions using a single technique at a time, followed by a combination of all techniques in the final stage. The AI-generated paraphrases underwent proofreading, and the effective ones were preserved in an answer keys document for students’ reference. Regarding the development of materials for summarizing skills, I employed XIPU AI to summarize the given passage within different word limits, e.g. fifty words, twenty-five words, and fifteen words. All the AI-generated summaries were systematically collected and saved in a Word document for subsequent use. Following traditional reading comprehension activities, students were tasked with summarizing the main ideas in their own words, adhering to specified word limits. Afterwards, they compared their summaries with the AI-generated versions and were encouraged to identify the most important words, eliminate redundant information, and study the paraphrasing or summarizing techniques applied. Moreover, when teaching students to use SWOT or PEST analytical tools for evaluating business propositions, XIPU AI played a demonstrative role. After introducing the SWOT or PEST matrix, XIPU AI was used to demonstrate how the SWOT or PEST analysis paragraphs could be summarized in bullet points and shown in matrixes (See figures below).
The third application of XIPU AI in material development and adaptation pertains to listening comprehension, note-taking, and vocabulary expansion. I used XIPU AI to generate note-completion exercises for listening transcripts, guiding students in practicing their note-taking skills and summarizing the main ideas of the listening materials. Before the transcripts, explicit instructions were provided to the AI, directing it to formulate notes in bullet points or phrases, emphasizing the utilization of paraphrasing and summarizing skills in note-taking. Upon receiving the generated results, I looked for the lexical and syntactic changes in the note-completion exercises and further designed some gap-fill or note-completion activities. As for helping students with vocabulary expansion, I was able to identify a list of important words. This involved specifying parameters for the AI to generate a vocabulary list that contains part of speech, phonetic transcriptions, English definitions, and sentence examples based on the adapted versions of the curriculum materials. The resultant vocabulary lists facilitated the creation of diverse exercises, including vocabulary-definition matching and sentence completion exercises, aiding students in broadening their lexical repertoire. Through such adaptations, vocabulary, collocations, and sentence structures under the same topics can be presented to students in different forms (Koraishi, 2023), which is an effective way to maximally exploit the language input resources, affording students increased exposure opportunities and mitigating potential boredom that students might have.
Evaluation of the effectiveness of AI-generated materials
The AI-generated materials prove instrumental in expediting lesson preparation. Notably, the tool significantly reduces the time required for crafting answer keys for paraphrasing exercises, generating vocabulary exercises, and summarizing the central ideas of written texts. In other words, in terms of instructional material preparation, XIPU AI emerges as a highly advantageous resource.
The utilization of the AI tool translates into tangible benefits, allowing for the provision of enhanced and high-quality comprehensible input to students. This efficiency imparts a heightened sense of preparedness, enabling instructors to engage students with heightened confidence and readiness. The resultant materials contribute to an enriched learning experience, fostering an environment conducive to effective language acquisition.
From the students' perspective, the worksheet materials developed or adapted with the assistance of AI are accessible on the Learning Mall platform. These materials serve a multifaceted purpose, aiding in the expansion of students' vocabulary base, reinforcement of key information retention, and refinement of notetaking, paraphrasing, and summarizing skills. The scaffolded exercises helped students build up their confidence and facilitated their language acquisition process, supporting them to become more skilful English users in their academic journey.
Discussion on the benefits and drawbacks
Based on the evaluation of the effectiveness of AI-generated materials, it can be seen that both educators and students can benefit from AI applications in language teaching and learning. Specifically, AI not only enabled teachers to improve their work efficiency and offered students with language learning resources in a variety of forms but also facilitated students’ language acquisition process and empowered students with more confidence in enhancing their academic writing skills.
However, it is imperative to acknowledge that the integration of AI does not obviate the need for the teacher's discernment and expertise. Throughout the material development and adaptation phases facilitated by AI, my own knowledge remains pivotal in proofreading, evaluating material quality, and effecting necessary revisions. Although I do believe that AI tools will be more and more popular among teaching practitioners as well as university students, it is also worthwhile for us to be aware of the potential risks of overdependence on AI (Kavanagh, 2022; Alammar and Amin, 2023). While enjoying the quick responses and immediate satisfaction of ready-made materials, we shall also be mindful of the diminished initiative, the lack of creativity (Alammar and Amin, 2023), and the growing anxiety in instances where AI is unavailable. As we rely increasingly on the applications of AI, both educators and students must maintain the ability to evaluate the quality and accuracy of AI-generated texts as well as uphold academic integrity policies (Alammar and Amin, 2023) to avoid misconduct or offenses.
Conclusion
This article shares my experience with using XIPU AI for material development and adaptations in language teaching. The practical applications include adjusting language input materials, designing paraphrasing and summarizing exercises, and creating note-completion and vocabulary expansion exercises. The XIPU AI generated materials were considered to be effective in terms of expediting lesson preparation and diversifying language learning activities. Although teachers do benefit from the integration of AI in language teaching, the discernment and expertise of teachers are still of paramount importance. Moreover, potential risks of an excessive reliance on AI should be taken caution with.
Reference
Alammar, A. and Amin, E. A. (2023) ‘EFL Students’ Perception of Using AI Paraphrasing Tools in English Language Research Projects’, Arab World English Journal, 14(3), pp. 166–181. doi:10.24093/awej/vol14no3.11.
Kavanagh, B. (2022) ‘Assessing AI-based Summarizing and Paraphrasing Tools for a CLIL Intercultural Communication Academic Writing Class’, Intercultural Communication Studies, 31(2), pp. 93–110. Available at: https://search-ebscohost-com-s.elink.xjtlu.edu.cn:443/login.aspxdirect=true&db=edo&AN=163900821&site=eds-live&scope=site (Accessed: 6 January 2024).
Koraishi, O. (2023) ‘Teaching English in the Age of AI: Embracing ChatGPT to Optimize EFL Materials and Assessment’, Language Education & Technology (LET Journal), 3(1), pp. 55–72. Available at: https://search-ebscohost-com-s.elink.xjtlu.edu.cn:443/login.aspx?direct=true&db=mlf&AN=EIS164791622&site=eds-live&scope=site (Accessed: 6 January 2024).
Regala-Flores, E. and Lopez, M. (2019) ‘Self-reported summarizing and paraphrasing difficulties in L2 writing contexts: Some pedagogical interventions’, Indonesian Journal of Applied Linguistics, 9(2), pp. 286–296. doi:10.17509/ijal.v9i2.20219.
