ABSTRACT

Professional translation was one of the first computational settings in which human users were paired with intelligent machine agents. During the first machine translation (MT) experiments in the early 1950s, it was observed that, to achieve a high level of final translation quality, MT output could aid human users, and that human users could provide valuable feedback to machines in the form of corrections. In this chapter, we review the advances in both the understanding and implementation of this human–machine interface. We review and discuss the use of human–computer interaction (HCI) methods in designing and evaluating translation aids, focusing specifically on interactive machine translation (IMT).