Usage of Genetic Algorithms for Educational Tests Adaptation

摘要

This paper surveys the web-based educational testing system consisting of four components that intended for use in several semi-automatic steps from assessment topics data filling to examination. The first two components are about filling the knowledge base of specific format and automatic generating of questions using this knowledge. Probability based branching tests building is provided by the third component and an examination with optional test adaptation are available through the fourth one. Test adaptation approach is based on slightly modified genetic algorithm which involves specific definition of entity genotype and it's relation to test data.