Can Machine Translation Be as Effective as a Human Translation? A Cross Linguistic Analysis of Machine Translation Ambiguity between English, French and Armenian.

Abstract

Statement/Research Question: Machine Translation (MT) still remains a tough challenge for both linguists and programmers. In spite of all the promises and hopes, it failed to meet the satisfactory standards. Actually, translation itself is a tough process, even for human beings. Can MT be compared with Human Translation? The paper discusses the difference between MT and Human Translation focusing on the effectiveness of each and pointing out semantic ambiguity in English, French and Armenian translations. Ambiguity still results in huge barriers on the effectiveness of Machine Translation. Methodology: We carried out an experiment by translating many texts of English, French and Armenian through different Machine Translators and analysed the error patterns1 by a dictionary method technique. One word can represent more than one meaning in a language. In turn, a word with identical meanings to the first language might have complete different meaning in the later. E.g. The words table (English), table (French) and սեղան (Armenian) express the same concept only in their general meaning, which is a piece of furniture. While the second meaning for the word table in English and French means a chart whereas in Armenian it is a trapezium. Findings: Finally, this paper also discusses the factors causing ambiguity in MT from two main perspectives: a) A cross-linguistics difference in second meaning of the words among these three languages. b) type/parts of speech causing more ambiguity over the others.



Author Information
Gohar Ghalachyan, Yerevan State University, Armenia

Paper Information
Conference: ECLL2015
Stream: Linguistics

This paper is part of the ECLL2015 Conference Proceedings (View)
Full Paper
View / Download the full paper in a new tab/window


Comments & Feedback

Place a comment using your LinkedIn profile

Comments

Share on activity feed

Powered by WP LinkPress

Share this Research

Posted by James Alexander Gordon