• Ni Nyoman Parmithi

  Author

• Ni Putu Sri Ratna Dewi

  Author

In the field of second or foreign language learning, Intelligent Tutoring Systems (ITS) based on Artificial Intelligence (AI) have been revolutionizing the approach to teaching and learning English by providing adaptive support, personalized instruction, and immediate feedback to learners. In the realm of second or foreign language learning, Intelligent Tutoring Systems (ITS) fueled by Artificial Intelligence (AI) are transforming the way educators teach and students learn English, by adapting instruction to align with the unique needs of each learner and providing immediate feedback. This paper presents the design, implementation, and classroom evaluation of an Intelligent Tutoring System based on Artificial Intelligence for English teaching that focuses on correcting grammar, learning vocabulary, and providing pronunciation feedback. A mixed-method design using prototype development and pilot classroom evaluation was used, with forty-five intermediate-level English learners engaging in the proposed system over six weeks, and the conventional system used for the comparison group. The data collected included pre- and post-test scores, questionnaires, and short interviews, as well as the system usage logs, and were analyzed using descriptive statistics and thematic coding of the learners' responses. The results showed significantly greater improvements in grammatical accuracy, vocabulary recall, and self-reported motivation for students using the AI tutoring system compared to the control group. The incorporation of NLP elements—such as automatic error detection, learner modeling, and generation of adaptive feedback—significantly shortened the delay between learner errors and corrective feedback, maintained learner involvement, and maintained the delivery of feedback. The study also highlights key design factors in determining the effectiveness of Intelligent Tutoring Systems in language learning: learner modeling, pedagogical strategy selection, and affect-aware feedback. Issues with data privacy, teacher preparedness, algorithmic bias and the ability of AI models to adapt to different linguistic contexts are also explored. The paper points out that AI-based tutoring systems have the potential to offer significant benefits for personalized English language learning at scale, but also emphasizes the importance of retaining human pedagogical control, continued assessment, and system design that is culturally responsive in the future.

• PDF