Impasse-Driven Evaluation and Assistance to Web Learners

摘要

Computer mediated educational systems have abilities to support both of students and a teacher in their activity of learning, teaching and collaboration. Basically web-based Interactive Learning Environments (ILE) provide students with unlimited access to teaching materials asynchronously. As a consequence, the teacher's load of educational efforts in web-based ILE education increases as compared with traditional classroom education. The difficult problem of designing web-based ILE systems is to estimate about what a student is learning, how the system supports a student to proceed and when a system interrupts student's learning. Some of innovative researches are pursuing the solution by evaluating the access records to web text pages as the student model . We are developing HyperClassroom which is a web-based educational system and serves students with three learning environments of interacting customized courseware delivered over World Wide Web (WWW), collaboration with staffs and/or students and an intelligent tutoring system. The student model tracks web learner's access records. The central function for estimating the student model is a fuzzy inference engine which evaluates the history of student's records which web text pages has been opened and how long each of the pages has been opened. A student does not always become aware of impasses in his or her learning processes. In our proposal, we categorize the impasses in three states, an impasse noticed by a student by himself or by herself, an impasse perceived by a teacher and an impasse reasoned by the estimated student model. When a teacher identifies one of the categorized impasses, the teacher uses the learning history of the target student for supporting the te! acher's pedagogical decision to choose more appropriate learning environment. For this purpose, HyperClassroom can make a communication channel between the student and the teacher. In this paper, we describe the concept of HyperClassroom and the method of student modeling at first. Then we show how HyperClassroom finds the impasse of a student during his or her learning and how the communication channel is made between a student and a teacher. Finally some experimental results of fuzzy inference engine are compared with the student's scores of final examination.