Target-Side Context for Discriminative Models in Statistical Machine Translation

TL;DR

This paper introduces a novel target-side context model for statistical machine translation that improves translation quality by better capturing morphological coherence, efficiently integrated into decoding, and scalable to large datasets.

Contribution

It extends discriminative translation models by incorporating target context, enhancing translation quality and morphological coherence in a scalable, efficient manner.

Findings

Consistent improvements across four language pairs.

Efficient integration into decoding process.

Enhanced morphological coherence capture.

Abstract

Discriminative translation models utilizing source context have been shown to help statistical machine translation performance. We propose a novel extension of this work using target context information. Surprisingly, we show that this model can be efficiently integrated directly in the decoding process. Our approach scales to large training data sizes and results in consistent improvements in translation quality on four language pairs. We also provide an analysis comparing the strengths of the baseline source-context model with our extended source-context and target-context model and we show that our extension allows us to better capture morphological coherence. Our work is freely available as part of Moses.