Fuzzy rules for tests complexity changing for individual learning path construction

Abstract

The fuzzy rules for changing of complexity level at adaptive testing were designed. Using of fussy rules for learning objects optimal selection for generation of individual learning path is proposed. Since the limits of student's knowledge assessment is difficult to determine, it is proposed to use the fuzzy approach. There is proved that using of fuzzy approach simplifies the evaluation of student's knowledge, because fuzzy systems are better understood by students and mark for answering test questions can be also considered as fuzzy. Fuzzy models are transparent enough and understandable, therefore they are acceptable when informative interpretation is more important than simulation accuracy. Experimental results show that adaptive test with fuzzy rules is three times faster that classical test.