Adaptive fuzzy ontology for student assessment

The traditional test usually uses a score to present the students´ learning performance; however, it seems difficult to clearly understand the students´ learning performance only by the score. As a result, this paper proposes an adaptive fuzzy ontology for student learning assessment and applies it to mathematics area. First, the domain experts construct the adaptive mathematics fuzzy ontology by referring to the guidelines of mathematics learning area in Grades 1-9 curriculum. The natural language processing mechanism tags each term with its speech and then filters the terms with useless speeches from the response data. Based on the genetic learning mechanism, the fuzzy reasoning mechanism then reasons the similarity strength between the kept terms and the constructed ontology. The semantic summary mechanism next summarizes the students´ learning performance based on the inferred results. Finally, the diagnosis report mechanism presents the diagnosed reports to make officers, teachers, and students themselves much understand examinees´ learning progress. Experimental results indicate that the proposed method can generate the suitable summarized sentences to allow teachers to quickly understand which mathematical topic is the one that students should be improved in the future.