An intelligent computer- based tutoring approach for the management of negative transfer

Abstract

This research addresses how a prototype of a language tutoring system, the Chinese Tutor, tackles the practical problem of negative transfer (i.e. mother tongue influence) in the learning of Chinese grammar by English-speaking students. The design of the Chinese Tutor has been based on the results of empirical studies carried out as part of this research. The results of the data analysis show that negative transfer can be used to account for almost 80% of the errors observed in the linguistic output of students in their study of Chinese. If the students can be helped to overcome these errors, the standard of their Chinese will be greatly improved. In this research, an approach of Intelligent Language Tutoring Systems (ILTSs) has been adopted for handling negative transfer. This is because there are several advantages of ILTSs, including interactive learning, highly individualised instruction and student-centred instruction [Wyatt 1984 .The Chinese Tutor contains five main components: the Expert Model, which contains all the linguistic knowledge for tutoring and serves as a standard for evaluating the student's performance; the Student Model, which collects information on the student's performance; the Diagnoser, which detects different types of error made by the student; the Tutor Model, which plans student learning, makes didactic decisions and chooses an appropriate tutorial strategy based on the student’s performance; and the Interface Module, which communicates between the student and the system. A general and robust solution to the treatment of negative transfer, i.e. the technique of Mixed Grammar has been devised. The rules in this grammar can be applied to detect arbitrary transfer errors by using a general set of rules. A number of students in the Department of East Asian Studies at the University of Durham have used the Chinese Tutor with positive results.