Ant Colony Optimization for IRT Based Test-sheet Composition

• Main Authors: Ang-Hua Li, 李昂樺
• Other Authors: Tsu-Feng Ho
• Format: Others
• Language: zh-TW
• Published: 2010
• Online Access: http://ndltd.ncl.edu.tw/handle/19167871141834055654

碩士 === 銘傳大學 === 資訊工程學系碩士班 === 98 === In order to precisely measure students’ ability, many efforts have been made to assemble test-sheet based on Item Response Theory (IRT), which is able to choose test items according to students’ ability levels. As the number of items of a test bank is getting larger, the number of test-sheet combinations increases explosively. In fact, problem of test sheet composition can be transformed to a combinational optimization problem and has been proven as a NP-Hard problem, which traditional algorithms are not able to find an optimal solution in an acceptable time. Many research works have focused on using heuristic algorithms to compose a test-sheet under all constraints in a reasonable time. In this proposal, currently, we developed a method to adjust neighborhood size for improving the effectiveness of Ant Colony Optimization. According to the experimental results, the proposed method can find a better solution in shorter time. In order to enhance the effectiveness of test-sheet combinations, the multi-dimensional IRT will be applied in the study.