Abstract:While NMT has achieved remarkable results in the last 5 years, production systems come with strict quality requirements in arbitrarily niche domains that are not always adequately covered by readily available parallel corpora. This is typically addressed by training domain specific models, using fine-tuning methods and some variation of back-translation on top of in-domain monolingual corpora. However, industrial practitioners can rarely afford to focus on a single domain. A far more typical scenario includes a set of closely related, yet succinctly different sub-domains. At <a class="link-external link-http" href="http://Booking.com" rel="external noopener nofollow">this http URL</a>, we need to translate property descriptions, user reviews, as well as messages, (for example those sent between a customer and an agent or property manager). An editor might need to translate articles across a set of different topics. An e-commerce platform would typically need to translate both the description of each item and the user generated content related to them. To this end, we propose MDT: a novel method to simultaneously fine-tune on several sub-domains by passing multidimensional sentence-level information to the model during training and inference. We show that MDT achieves results competitive to N specialist models each fine-tuned on a single constituent domain, while effectively serving all N sub-domains, therefore cutting development and maintenance costs by the same factor. Besides BLEU (industry standard automatic evaluation metric known to only weakly correlate with human judgement) we also report rigorous human evaluation results for all models and sub-domains as well as specific examples that better contextualise the performance of each model in terms of adequacy and fluency. To facilitate further research, we plan to make the code available upon acceptance.

General2Specialized LLMs Translation for E-commerce

Llms-Based Machine Translation for E-Commerce

Bilingual Terminology Extraction from Comparable E-Commerce Corpora

Investigating LLM Applications in E-Commerce

EcomGPT: Instruction-tuning Large Language Models with Chain-of-Task Tasks for E-commerce

eCeLLM: Generalizing Large Language Models for E-commerce from Large-scale, High-quality Instruction Data

EcomGPT-CT: Continual Pre-training of E-commerce Large Language Models with Semi-structured Data

Multilingual Machine Translation with Large Language Models: Empirical Results and Analysis

Enhancing E-commerce Product Title Translation with Retrieval-Augmented Generation and Large Language Models

On the Principles and Decisions of New Word Translation in Sino-Japan Cross-Border e-Commerce: A Study in the Context of Cross-Cultural Communication

EC-Guide: A Comprehensive E-Commerce Guide for Instruction Tuning and Quantization

Multi-Domain Adaptation in Neural Machine Translation Through Multidimensional Tagging

Document-Level Machine Translation with Large Language Models

Cross-Lingual Low-Resource Set-to-Description Retrieval for Global E-Commerce

Exploring the traditional NMT model and Large Language Model for chat translation

(Perhaps) Beyond Human Translation: Harnessing Multi-Agent Collaboration for Translating Ultra-Long Literary Texts

Unified Model Learning for Various Neural Machine Translation

Guided Alignment Training for Topic-Aware Neural Machine Translation

Multi-Domain Neural Machine Translation with Word-Level Domain Context Discrimination

Learning to Generalize to More: Continuous Semantic Augmentation for Neural Machine Translation

Towards Reliable Neural Machine Translation with Consistency-Aware Meta-Learning