An English Dialogue Companion System for Supporting Conversation Practice

In this paper, we proposed an English dialogue companion system, called EDC, using predicate argument structure by Support Vector Machine (SVM) classifier for semantic analyzer in order to support elementary students for learning English one on one. Three phases of learning activities were designed to enhance the English learning environment, which are the choosing learning companion phase, the conversation phase, and the teaching phase. We also conducted a between-subject design experiment to evaluate the effectiveness of the EDC system in terms of English achievement and using experience. Results of this study showed that there is no significant difference between the experimental group and the control group on English achievement. However, more than half subjects in the lower score group got higher score in post-test than in pre-test. Thus, we assumed that the system could be useful for students in the lower English achievement degree.