One aim of testing is to identify weaknesses in students' knowledge. Computerized tests are now one of the most important ways to judge learning, and, selecting tailored questions for each learner is a significant part of such tests. Therefore, one current trend is that computerized adaptive tests (CATs) not only assist teachers in estimating the learning performance of students, but also facilitate understanding of problems in their learning process. These tests, must effectively and efficiently select questions from a large-scale item bank, and to cope with this problem we propose a dynamic question generation system for web-based tests using the novel approach of particle swarm optimization (PSO). The dynamic question generation system is built to select tailored questions for each learner from the item bank to satisfy multiple assessment requirements. Furthermore, the proposed approach is able to efficiently generate near-optimal questions that satisfy multiple assessment criteria. With a series of experiments, we compare the efficiency and efficacy of the PSO approach with other approaches. The experimental results show that the PSO approach is suitable for the selection of near-optimal questions from large-scale item banks.
All Science Journal Classification (ASJC) codes
- Computer Science Applications
- Artificial Intelligence