Automatic question generation system for learning to create linear programming models

Abstract

Automatic question generation systems play a key role in enhancing the efficiency of teaching and learning, particularly in fields that involve complex problem-solving, such as linear programming (LP). This study presents the development and evaluation of a system designed to generate questions and answers related to business product mix problems in LP. Aimed at enhancing LP modeling skills, the system was tested on 132 undergraduate business students enrolled in a quantitative analysis course. The evaluation involved pre-and post-learning achievement tests, with data analyzed using t-tests and normalized gain (g) to measure learning progress. Results showed significant improvements in students’ performance on identical (t = 14.94, p<0.05) and different tests (t = 8.95), along with a moderate learning gain (g = 0.59). These findings indicate that the system not only reduces the instructional workload but also effectively enhances students’ understanding and application of LP concepts, making it a valuable tool in education. © 2025, Silpakorn University. All rights reserved.