User-side assessment on English to Thai and Thai to English machine interpreting: the case of google translate

Abstract

This research aims to assess Google Translate´s effectiveness in English-Thai consecutive interpreting based on users' comprehension and to investigate users’ satisfaction with MI renditions. Ten native speakers of Thai and eight native speakers of English were asked to listen to the translated speeches and complete a questionnaire to assess the Adequacy and Fluency of the MI. The findings showed that in terms of Adequacy of the English to Thai MI participants’ comprehension scores ranged from 2.4 to 3.45 out of 5, while Fluency received less than half of the maximum scores in all areas. In comparison, for the direction of Thai to English of the MI, the comprehension scores of the participants ranged from 2.31 and 3.81 out of 5 for Adequacy, and three areas of Fluency received less than half of the maximum scores. However, improvements are still needed. For the English to Thai direction, appropriate pausing of translation delivery was the most suggested aspect. On the other hand, a more natural-sounding voice is mostly required in the Thai to English direction.