Mixing Words and Emotions - New Natural Methods for Artificial Language Tutors

Abstract

In this paper we would like to present the idea of an English tutoring system aimed to bring a new standard in language teaching. The Co-Mix project concentrates on second language vocabulary acquisition that is central to language itself and of high importance to the typical language learner. We compare some existing methods of teaching vocabulary and present the usefulness of our sentential context-learning method based on the phenomenon of code-mixing. Since such a method has not yet been fully explored, we would like to progress with the research in the fields of natural language processing and artificial tutor systems. We also perform a preliminary experiment to check our hypothesis on sentences with certain emotional loads. We investigate the effect of emotion on the learner's performance in studying and propose how this method can be applied in a casual conversation system with the purpose of teaching English as a second language.