Student Evaluation Based On Fuzzy Rule Interpolation

Abstract

The evaluation of narrative student responses contains always vague elements that originate from the fact that the students performance cannot be clearly ranked in one or another category, which makes this area a natural application field for fuzzy set theory based solutions. This paper presents a new evaluation method that allows taking into consideration three main aspects, i.e. the accuracy of the response, the correctness of the technical terms’ use, and the total time necessary for answering the questions. The evaluator can express the uncertainty in his or her judgment by fuzzy numbers. The method determines the final evaluation using fuzzy arithmetic and inference based on rule interpolation.