Quality Estimation of Machine Translated Texts based on Direct Evidence from Training Data

TL;DR

This paper proposes a novel approach to estimate machine translation quality by leveraging direct clues from the training data, demonstrating its effectiveness across data-driven systems.

Contribution

It introduces a simple method that uses training corpus information to assess translation quality without reference translations, advancing quality estimation techniques.

Findings

Training data contains direct clues for quality estimation

Method is effective across various data-driven MT systems

Approach does not require reference translations

Abstract

Current Machine Translation systems achieve very good results on a growing variety of language pairs and data sets. However, it is now well known that they produce fluent translation outputs that often can contain important meaning errors. Quality Estimation task deals with the estimation of quality of translations produced by a Machine Translation system without depending on Reference Translations. A number of approaches have been suggested over the years. In this paper we show that the parallel corpus used as training data for training the MT system holds direct clues for estimating the quality of translations produced by the MT system. Our experiments show that this simple and direct method holds promise for quality estimation of translations produced by any purely data driven machine translation system.