A High-Quality Multilingual Dataset for Structured Documentation Translation

TL;DR

This paper introduces a high-quality multilingual dataset of XML-structured documentation for translation, enabling improved translation models that preserve structure and support multiple language pairs, including non-English ones.

Contribution

The paper provides a novel XML-structured parallel dataset for documentation translation and evaluates models that incorporate XML constraints and copy mechanisms for better accuracy.

Findings

XML-aware translation improves accuracy

Beam search effectively maintains XML structure

Copy mechanisms impact translation of numerical data

Abstract

This paper presents a high-quality multilingual dataset for the documentation domain to advance research on localization of structured text. Unlike widely-used datasets for translation of plain text, we collect XML-structured parallel text segments from the online documentation for an enterprise software platform. These Web pages have been professionally translated from English into 16 languages and maintained by domain experts, and around 100,000 text segments are available for each language pair. We build and evaluate translation models for seven target languages from English, with several different copy mechanisms and an XML-constrained beam search. We also experiment with a non-English pair to show that our dataset has the potential to explicitly enable $17 \times 16$ translation settings. Our experiments show that learning to translate with the XML tags improves translation accuracy, and the beam search accurately generates XML structures. We also discuss trade-offs of using the copy mechanisms by focusing on translation of numerical words and named entities. We further provide a detailed human analysis of gaps between the model output and human translations for real-world applications, including suitability for post-editing.