Translation Transformers Rediscover Inherent Data Domains

TL;DR

This paper reveals that NMT Transformers inherently encode domain information in sentence representations, enabling effective unsupervised domain clustering and adaptation without external models, outperforming pre-trained LMs.

Contribution

It demonstrates that NMT models internally encode domain data, allowing unsupervised clustering and domain adaptation, which surpasses previous reliance on external language models.

Findings

NMT sentence representations encode domain information explicitly.

NMT clustering aligns closely with actual domains, especially at document level.

Using NMT for domain clustering outperforms pre-trained language models.

Abstract

Many works proposed methods to improve the performance of Neural Machine Translation (NMT) models in a domain/multi-domain adaptation scenario. However, an understanding of how NMT baselines represent text domain information internally is still lacking. Here we analyze the sentence representations learned by NMT Transformers and show that these explicitly include the information on text domains, even after only seeing the input sentences without domains labels. Furthermore, we show that this internal information is enough to cluster sentences by their underlying domains without supervision. We show that NMT models produce clusters better aligned to the actual domains compared to pre-trained language models (LMs). Notably, when computed on document-level, NMT cluster-to-domain correspondence nears 100%. We use these findings together with an approach to NMT domain adaptation using automatically extracted domains. Whereas previous work relied on external LMs for text clustering, we propose re-using the NMT model as a source of unsupervised clusters. We perform an extensive experimental study comparing two approaches across two data scenarios, three language pairs, and both sentence-level and document-level clustering, showing equal or significantly superior performance compared to LMs.