Toward a science of machine translation

The fact that machine translation output does not reach the level of good human translations is well known. However, there has been surprisingly little attention paid by the machine translation community to how humans achieve such results. In this paper,...

Full description

Saved in:
Bibliographic Details
Main Author: Bond, Francis.
Other Authors: School of Humanities and Social Sciences
Format: Conference or Workshop Item
Language:English
Published: 2011
Subjects:
Online Access:https://hdl.handle.net/10356/93796
http://hdl.handle.net/10220/7273
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
Description
Summary:The fact that machine translation output does not reach the level of good human translations is well known. However, there has been surprisingly little attention paid by the machine translation community to how humans achieve such results. In this paper, I suggest several ways to improve machine translation, based on the best practices of human translators, as described in Nida's (1964) Toward a Science of Translating. I call this approach multi-pass machine translation (MPMT), as it crucially relies on processing the text more than once. It is similar to the opportunistic bricoleur approach of Gdaniec (1999) in that it sets out to use the means at hand, adding to or changing them as necessary. As Sch utz (2001) points out, much of the research in the past decade has concentrated on the important but non-core issues of integrating MT into DTP formats and HTML. In this paper I concentrate on improving the MT engine itself. The resulting approach integrates much recent research into a single system.