Work

An AI-Based Commentary Function with Estimated State Rating Framework for Student Learning Support in Java Programming Learning Assistant System

Year: 2026
Type: article
Abstract: As Java remains a widely used object-oriented programming language in various application systems, we have developed the Java Programming Learning Assistant System (JPLAS) to support novice learners in their self-studies. JPLAS provides a variety of exercise problems that have different learning goals and difficulties. Among them, Grammar-Concept Understanding Problems (GUP) is designed as the first step problem where key grammar concepts appearing in a given source code are questioned. In this paper, we propose a commentary function supported by generative AI for GUP to study challenging concepts of Java programming such as generics and lambda expressions. This function provides AI-generated four-level hints to help students solve given questions. Besides, the Estimated State Rating (ESR) framework is presented to numerically describe the learning stagnation of a student. ESR is calculated using the time interval elapsed from the previous hint access and the current answer accuracy. For evaluations, we assigned the 10 GUP instances with the commentary function to 33 students in Okayama University, Japan, and calculated ESR from their log data. The evaluation examines the ESR values associated with the first recorded access to each commentary function and derives the corresponding 75th percentile as a reasonable reference threshold to allow the next-level hint. In future work, we will apply the proposal to other students and examine the effects on their problem solving and learning outcomes.
Source: Electronics
Institution: Okayama University
Language: English
Open Access status: gold
Estimated APC paid: $2,896
OpenAlex ID: W7226028319
FWCI:
Cites: 14
Cited by:
Related to: 10