المستخلص: |
The fact that completely programmed translation in numerous sets is away from creating an equivalent or prevalent yield to human translation has led to a serious enthusiasm for translation assessment in the MT people group. Nonetheless, investigation in this field, at this point, has not just overlooked the enormous measures of applicable learning accessible in a firmly related discipline to a great extent, specifically translation ponders; however it also neglected to give a more profound comprehension of the idea of "translation blunders" and "translation quality". The current paper represents a clarification of the quality notion in translation, by contrasting automatic and human assessments of students’ translations in the KOPTE selected corpus. It also demonstrates that studies of translation give modern detailed ideas for translation quality estimation and mistake comment. Moreover, by applying entrenched MT assessment scores, BLUE and Meteor, across KOPTE excerpts which were reviewed by a human professional, the recent research is keen to reveal insight into qualities (possible deficiencies) of such grades.
|