Exploiting Domain-Specific Parallel Data on Multilingual Language Models for Low-resource Language Translation
This work addresses the challenge of building effective neural machine translation systems for low-resource languages in specific domains, but it appears incremental as it builds on existing multilingual language models.
The paper tackles the problem of low-resource language translation in domain-specific settings by evaluating techniques like fine-tuning and further pre-training with auxiliary parallel data, finding that these strategies can improve model performance, though specific numerical gains are not detailed.
Neural Machine Translation (NMT) systems built on multilingual sequence-to-sequence Language Models (msLMs) fail to deliver expected results when the amount of parallel data for a language, as well as the language's representation in the model are limited. This restricts the capabilities of domain-specific NMT systems for low-resource languages (LRLs). As a solution, parallel data from auxiliary domains can be used either to fine-tune or to further pre-train the msLM. We present an evaluation of the effectiveness of these two techniques in the context of domain-specific LRL-NMT. We also explore the impact of domain divergence on NMT model performance. We recommend several strategies for utilizing auxiliary parallel data in building domain-specific NMT models for LRLs.