Combination of fuzzy and cognitive theories for adaptive e-assessment

Abstract

A crucial factor for successful educational results in computer-based educational systems and e-learning systems is the learner’s assessment. Assessment is more effective when it is tailored to each individual student’s learning needs and abilities. Therefore, a significant research challenge is to create tests that include exercises/questions/activities etc., which conform to each learner’s knowledge level and learning needs and abilities. This goal constitutes the need for creating adaptive tests. However, the area of adaptive e-assessment has not yet been explored sufficiently and thus there is scope for a lot of improvement. To this end, in this paper we present a novel solution for adaptive e-assessment. The novelty and significance lie in the blending of fuzzy logic and cognitive theories for further enhancing the personalization and adaptivity in e-assessment. Particularly, fuzzy sets are used to describe the knowledge level of students in a more realistic way. Furthermore, the cognitive theory of Revised Bloom Taxonomy is used to express the learning objectives that are required to be assessed through the created test. In addition, a fuzzy rule-based reasoner, which decides about the number and the difficulty level of the test items that have to be included into the created personalized test for each level of the Revised Bloom Taxonomy, is used. The fuzzy rules are applied to the fuzzy sets that describe the learners’ knowledge level. For the formation of the fuzzy sets and rules, the opinion of several tutors, holding experience in the educational process and instruction, was taken into consideration. The created adaptive test comprises distinct test items based on the individual learning needs of each student. The presented method has been used in two tutoring systems and has been fully evaluated. The evaluation results show great accuracy in the selection of test items for each individual student.