Neural and Statistical Methods for Leveraging Meta-information in Machine Translation

TL;DR

This paper explores neural and statistical techniques to incorporate meta-information, like text categories, into machine translation, resulting in up to 3% BLEU score improvements.

Contribution

It introduces neural network methods within a statistical machine translation framework to leverage meta-information for improved translation quality.

Findings

Up to 3% BLEU score improvement in certain categories

Neural methods effectively incorporate meta-information into SMT

Framework can be extended to various meta-data types

Abstract

In this paper, we discuss different methods which use meta information and richer context that may accompany source language input to improve machine translation quality. We focus on category information of input text as meta information, but the proposed methods can be extended to all textual and non-textual meta information that might be available for the input text or automatically predicted using the text content. The main novelty of this work is to use state-of-the-art neural network methods to tackle this problem within a statistical machine translation (SMT) framework. We observe translation quality improvements up to 3% in terms of BLEU score in some text categories.