کاربرد شبکه های پتری فازی سطح بالا برای سیستم درجه بندی آموزشی

For the purpose of understanding the students’ learning achievement, the most direct way is to implement a test. Due to the rapid development of information technology, all kinds of combination of information technology with the adaptive test have been incessantly noted by many scholars. In general, the computerized adaptive test includes the item response theory that tests the students’ learning ability of subjects. However, the results based only on the dichotomy of correct answers and wrong answers are not so comprehensive judgments. Situations of correct answers and wrong answers should be different in their degrees; for example, completely correct, partially correct, completely wrong, and partially wrong. But the partially correct or partially wrong is vague and difficult to define. Thus it is appropriate to use fuzzy theory to solve the vagueness problem. Therefore, this study presents a novel learning evaluation model which applies high-level fuzzy Petri net (HLFPN) and infers via a fuzzy reasoning method the different answering performances generated by different examinee’s abilities corresponding to the test items in different degrees of difficulty. Finally, we synthesize the answering performance of every test item and make a reasonable evaluation.