TY - GEN
T1 - A Case Study with an Intelligent Tutoring System for Active Study of Worked Examples in Introductory JAVA Programming
AU - Narayanan, Arun Balajiee Lekshmi
AU - Hassany, Mohammad
AU - Hendrawan, Rully
AU - Akhuseyinoglu, Kamil
AU - Brusilovsky, Peter
N1 - Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2026.
PY - 2026
Y1 - 2026
N2 - Worked examples are step-by-step solutions to problems in a specific domain, offered to students to acquire domain-specific problem-solving skills. We could magnify the power of worked examples by combining them with self-explanations, a strategy in which students explain each problem-solving step rather than passively studying it. In this work, we improve an existing system for automatically assessing student explanations to support the active study of worked examples. In particular, we use LLMs to address the limitations of semantic similarity methods for automatically assessing student explanations. We compare several methods based on semantic similarity with directly prompting LLMs. We showed that LLMs achieve an F1 score of 0.98, outperforming all semantic similarity baselines (F1 = 0.94). Our proposed work will improve the automated assessment and evaluate the system in a user study.
AB - Worked examples are step-by-step solutions to problems in a specific domain, offered to students to acquire domain-specific problem-solving skills. We could magnify the power of worked examples by combining them with self-explanations, a strategy in which students explain each problem-solving step rather than passively studying it. In this work, we improve an existing system for automatically assessing student explanations to support the active study of worked examples. In particular, we use LLMs to address the limitations of semantic similarity methods for automatically assessing student explanations. We compare several methods based on semantic similarity with directly prompting LLMs. We showed that LLMs achieve an F1 score of 0.98, outperforming all semantic similarity baselines (F1 = 0.94). Our proposed work will improve the automated assessment and evaluate the system in a user study.
KW - Evaluation
KW - Feedback
KW - LLMs
KW - Programming
KW - Self–Explanations
UR - https://www.scopus.com/pages/publications/105043921294
U2 - 10.1007/978-3-032-29794-5_71
DO - 10.1007/978-3-032-29794-5_71
M3 - Conference contribution
AN - SCOPUS:105043921294
SN - 9783032297938
T3 - Communications in Computer and Information Science
SP - 456
EP - 461
BT - Artificial Intelligence in Education. Late Breaking Results, WideAIED, Practitioners, Industry and Policies, Blue Sky, Doctoral Consortium, FoL Workshops and Tutorials, FoL Invited Papers - 27th International Conference, AIED 2026, Proceedings
A2 - Blanchard, Emmanuel G.
A2 - Chen, Guanliang
A2 - Chi, Min
A2 - Isotani, Seiji
PB - Springer Science and Business Media Deutschland GmbH
T2 - 27th International Conference on Artificial Intelligence in Education, AIED 2026
Y2 - 27 June 2026 through 3 July 2026
ER -