Nguyen, M. L., Hui, S. C. and Fong, A. C.M. (2011) A divide-and-conquer Tabu search approach for online test paper generation. In: 24th Australasian Joint Conference on Advances in Artificial Intelligence, Perth, Australia, 5-8 Dec 2011, pp. 717-726. (doi: 10.1007/978-3-642-25832-9_73)
Full text not currently available from Enlighten.
Publisher's URL: http://dx.doi.org/10.1007/978-3-642-25832-9_73
Abstract
Online Test Paper Generation (Online-TPG) is a promising approach for Web-based testing and intelligent tutoring. It generates a test paper automatically online according to user specification based on multiple assessment criteria, and the generated test paper can then be attempted over the Web by user for self-assessment. Online-TPG is challenging as it is a multi-objective optimization problem on constraint satisfaction that is NP-hard, and it is also required to satisfy the online runtime requirement. The current techniques such as dynamic programming, tabu search, swarm intelligence and biologically inspired algorithms are ineffective for Online-TPG as these techniques generally require long runtime for generating good quality test papers. In this paper, we propose an efficient approach, called DAC-TS, which is based on the principle of constraint-based divide-and-conquer (DAC) and tabu search (TS) for constraint decomposition and multi-objective optimization for Online-TPG. Our empirical performance results have shown that the proposed DAC-TS approach has outperformed other techniques in terms of runtime and paper quality.
Item Type: | Conference Proceedings |
---|---|
Additional Information: | <br>AI 2011: Advances in Artificial Intelligence</br> Lecture Notes in Computer Science; 7106 |
Status: | Published |
Refereed: | Yes |
Glasgow Author(s) Enlighten ID: | Fong, Dr Alvis Cheuk Min |
Authors: | Nguyen, M. L., Hui, S. C., and Fong, A. C.M. |
College/School: | College of Science and Engineering > School of Computing Science |
ISSN: | 0302-9743 |
University Staff: Request a correction | Enlighten Editors: Update this record