Analysing Quality of English-Hindi Machine Translation Engine Outputs Using Bayesian Classification

TL;DR

This paper proposes a Bayesian classification approach to evaluate the quality of English-Hindi machine translation outputs using 16 extracted features, aiming for more accurate sentence-level quality assessment.

Contribution

It introduces a novel Bayesian inference-based method utilizing 16 features for automatic translation quality estimation at the sentence level.

Findings

Bayesian model effectively predicts translation quality

Features improve correlation with human judgments

Method outperforms traditional metrics at segment level

Abstract

This paper considers the problem for estimating the quality of machine translation outputs which are independent of human intervention and are generally addressed using machine learning techniques.There are various measures through which a machine learns translations quality. Automatic Evaluation metrics produce good co-relation at corpus level but cannot produce the same results at the same segment or sentence level. In this paper 16 features are extracted from the input sentences and their translations and a quality score is obtained based on Bayesian inference produced from training data.