Application of particle swarm optimization to create multiple-choice tests

Generating tests from question banks by using manually extracted items or involving random method consumes a great deal of time and effort. At the same time, the quality of the resulting tests is often not high. The generated tests may not entirely meet the requirements formulated in advance. Therefore, this study develops innovative ways to enhance this process by optimizing the execution time and generating results that closely meet the extraction requirements. The paper proposes the use of Particle Swarm Optimization (PSO) to generate multiple-choice tests based on assumed objective levels of difficulty. The experimental results reveal that PSO speed-ups the extraction process, and improves the quality of tests in comparison with the results produced by previously used methods such as Random or Genetic Algorithm (GA) optimized methods. In addition, PSO shows to be more efficient than GA and random selection in most criteria, such as execution time, search space, stability, and standard deviation.