OPTIMISING EXPLAINABLE AI IN EDUCATION: A DSPY-BASED FRAMEWORK WITH CHAIN-OF-THOUGHT REASONING FOR ADAPTIVE LEARNING
DOI:
https://doi.org/10.33480/jitk.v12i1.8319Keywords:
Artificial Intelligence in Education, Chain-of-Thought Reasoning, DSPy, Explainable AI, Modular Learning SystemsAbstract
The application of artificial intelligence (AI) in education is often constrained by limited reasoning transparency, computational demands, and reproducibility challenges. This study proposes and conducts an exploratory evaluation of a modular AI education framework based on Declarative Structured Programming (DSPy) and Chain-of-Thought reasoning. The framework integrates typed input–output signatures with structured inference to support transparent question generation and adaptive feedback. A pilot study involved 10 participants with computer-science or education backgrounds; each completed three sessions, yielding 30 session-level observations. The framework used GPT-4o mini and was compared with prompt-based and rule-based baselines. It achieved 92.4% session-level learning accuracy and higher observed reasoning-clarity ratings than the baselines, with an average response time of 1.2 s. Because the sample was small, technically oriented, and evaluated over short sessions, the findings constitute preliminary evidence and should not be generalized to diverse learners or sustained learning outcomes. Larger, heterogeneous, longitudinal, and resource-instrumented studies are require.
Downloads
References
[1]. UNESCO, Guidance for Generative AI in Education and Research. Paris, France: UNESCO, 2023.
[2]. A. Tlili et al., “What if the devil is my guardian angel: ChatGPT as a case study of using chatbots in education,” Smart Learn. Environ., vol. 10, art. no. 15, 2023, doi: 10.1186/s40561-023-00237-x.
[3]. E. Kasneci et al., “ChatGPT for good? On opportunities and challenges of large language models for education,” Learn. Individ. Differ., vol. 103, art. no. 102274, 2023, doi: 10.1016/j.lindif.2023.102274.
[4]. W. Holmes and K. Porayska-Pomsta, Eds., The Ethics of Artificial Intelligence in Education: Practices, Challenges, and Debates. New York, NY, USA: Routledge, 2022. doi: https://doi.org/10.4324/9780429329067.
[5]. H. Khosravi et al., “Explainable artificial intelligence in education,” Comput. Educ. Artif. Intell., vol. 3, art. no. 100074, 2022, doi: 10.1016/j.caeai.2022.100074.
[6]. Y. K. Dwivedi et al., “So what if ChatGPT wrote it? Multidisciplinary perspectives on opportunities, challenges and implications of generative conversational AI for research, practice and policy,” Int. J. Inf. Manage., vol. 71, art. no. 102642, 2023, doi: 10.1016/j.ijinfomgt.2023.102642.
[7]. C. K. Lo, “What is the impact of ChatGPT on education? A rapid review of the literature,” Educ. Sci., vol. 13, no. 4, art. no. 410, 2023, doi: 10.3390/educsci13040410.
[8]. S. S. Gill et al., “Transformative effects of ChatGPT on modern education: Emerging era of AI chatbots,” Internet Things Cyber-Phys. Syst., vol. 4, pp. 19–23, 2024, doi: 10.1016/j.iotcps.2023.06.002.
[9]. S. J. Lee and K. K. Kwon, “A systematic review of AI education in K–12 classrooms from 2018 to 2023,” Comput. Educ. Artif. Intell., vol. 6, art. no. 100211, 2024, doi: 10.1016/j.caeai.2024.100211.
[10]. X. Wang et al., “Self-consistency improves chain-of-thought reasoning in language models,” in Proc. Int. Conf. Learn. Representations (ICLR), 2023, doi: 10.48550/arXiv.2203.11171.
[11]. D. Zhou et al., “Least-to-most prompting enables complex reasoning in large language models,” in Proc. Int. Conf. Learn. Representations (ICLR), 2023, doi: 10.48550/arXiv.2205.10625.
[12]. R. F. Kizilcec, “To advance AI use in education, focus on understanding educators,” Int. J. Artif. Intell. Educ., vol. 34, pp. 12–19, 2024, doi: 10.1007/s40593-023-00351-4.
[13]. J. Wei et al., “Chain-of-thought prompting elicits reasoning in large language models,” in Advances in Neural Information Processing Systems, vol. 35, 2022, pp. 24824–24837, doi: 10.48550/arXiv.2201.11903.
[14]. O. Khattab et al., “DSPy: Compiling declarative language model calls into self-improving pipelines,” in Proc. Int. Conf. Learn. Representations (ICLR), 2024, doi: 10.48550/arXiv.2310.03714.
[15]. B. Rahman and I. Sutedja, “AI based sagittal spinal posture assessment for adolescent screening in low resource school settings,” Sci. Rep., 2026, doi: 10.1038/s41598-026-36634-8.
[16]. B. Rahman, Maryani, and T. T., “IDChat: Toward a universal, multi-biometric digital identity for next-generation secure communication in ASEAN,” IEEE Access, vol. 14, pp. 14892–14902, 2026, doi: 10.1109/ACCESS.2025.3645071.
[17]. B. Rahman and Maryani, “Optimizing customer satisfaction through sentiment analysis: A BERT-based machine learning approach to extract insights,” IEEE Access, vol. 12, pp. 151476–151489, 2024, doi: 10.1109/ACCESS.2024.3478835.
[18]. B. Rahman, “Development of an image-based road pothole detection model with U-Net implementation,” in Proc. 2023 IEEE 9th Int. Conf. Computing, Engineering and Design (ICCED), 2023, pp. 1–6, doi: 10.1109/ICCED60214.2023.10425723.
[19]. A. Singhvi et al., “DSPy assertions: Computational constraints for self-refining language model pipelines,” arXiv:2312.13382, 2023, doi: 10.48550/arXiv.2312.13382.
[20]. A. Madaan et al., “Self-refine: Iterative refinement with self-feedback,” arXiv:2303.17651, 2023, doi: 10.48550/arXiv.2303.17651.
[21]. L. Longo et al., “Explainable artificial intelligence (XAI) 2.0: A manifesto of open challenges and interdisciplinary research directions,” Inf. Fusion, vol. 106, art. no. 102301, 2024, doi: 10.1016/j.inffus.2024.102301.
[22]. R. Alfredo et al., “Human-centred learning analytics and AI in education: A systematic literature review,” Comput. Educ. Artif. Intell., vol. 6, art. no. 100215, 2024, doi: 10.1016/j.caeai.2024.100215.
[23]. L. Zheng et al., “Judging LLM-as-a-judge with MT-Bench and Chatbot Arena,” in Advances in Neural Information Processing Systems, vol. 36, 2023, pp. 46595–46623, doi: 10.48550/arXiv.2306.05685.
[24]. D. L. Naryatmojo and B. Hartono, “The effectiveness of the project-based learning model assisted by artificial intelligence on the listening skills of Indonesian language education students,” JTP—J. Teknol. Pendidik., vol. 28, no. 1, pp. 100–112, 2026, doi: 10.21009/jtp.v28i1.65961.
[25]. Supriyadi and A. N. Amalia, “Artificial intelligence integration and students’ digital literacy as determinants of learning outcomes in an economics education assessment course,” J. Pendidik. Ekon. Bisnis, vol. 13, no. 1, pp. 118–126, 2026, doi: 10.21009/JPEB.013.1.8.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Ben Rahman
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
-a.jpg)
-b.jpg)
