Document Type

Project

Publication Date

12-15-2017

Keywords

Machine translation, Computer Science

Abstract

Chinese second language learners of English often use Machine Translators (MT) to translate personal and professional messages from their first language to English. MT’s are not perfect and have historically create messages that lack the cohesiveness and authenticity of natively written English. This paper describes our attempts to quantify the differences between human translation and machine translation in a specific scope with that hope that both MTs and post editing systems can be benefited through awareness of common error and differences between human and machine translations. In order to achieve this we implemented existing algorithms designed to identify common errors in machine translated sentences and a sentence dependency analysis to identify key differences between the translations. With this information further work targeting the underlying causes of these errors can be developed.

Language

English

Format

application/pdf

Share

COinS