ANFIS supported question classification in computer adaptive testing (CAT)

Abstract

E-Iearning has become a major trend in the computer assisted teaching with the rapid development of Internet technologies. Web-based education is a very important component of education technology. One of the main advantage is the classroom and platform independence. Implementing Artificial Intelligence (AI) techniques to support efforts to improve the Web's intelligence and provide better services to the end users. In this study, three popular AI methods: Artificial Neural Network (ANN), Support Vector Machines (SVM), and Adaptive Network Based Fuzzy Inference System (ANFIS) were benchmarked in terms of effectiveness and performance within a Web-based environment. As the pilot test, History of Civilization class was selected. The question classification abilities depending on the item responses of students, item difficulties of questions, and question levels were determined by using Gaussian Normal Curve. Comparison study was conducted by considering the performance and class correctness of the sample questions (n=13) by using the given three inputs. The results showed that ANFIS has better performance than ANN and AVM in web-based education. ©2009 IEEE. © 2010 Elsevier B.V., All rights reserved.