Testing students' understanding of complex transfer

Abstract

Courses on Machine Translation (MT) need to be tailored to different sets of students with differing skills and demands (Kenny & Way, 2001) Nevertheless, any contemporary course on MT ought to equip students with at least a superficial knowledge of the differences between rule-based and statistical MT direct and indirect approaches and transfer- based and interlingual systems. With regard to this latter distinction, the issue of complex transfer is an integral component to this section of a course on MT, whether this be to computational linguists, translators or language students. This paper presents a method of assessing the level of understanding of the issues pertaining to complex transfer for final year undergraduates studying a degree programme in Computational Linguistics. The intention is that this methodology may contribute to a suite of exercises which may be used by other instructors in Machine Translation.