Domain Adaptation in Neural Machine Translation using a Qualia-Enriched FrameNet
This addresses domain adaptation challenges for NMT, particularly benefiting lesser-resourced languages by avoiding fine-tuning, though it is incremental as it builds on existing FrameNet and qualia concepts.
The paper tackles domain adaptation in Neural Machine Translation by introducing Scylla, a method that uses a qualia-enriched FrameNet as an external knowledge base without fine-tuning, and it significantly outperforms a commercial baseline in HTER for translating sports domain sentences from Brazilian Portuguese to English.
In this paper we present Scylla, a methodology for domain adaptation of Neural Machine Translation (NMT) systems that make use of a multilingual FrameNet enriched with qualia relations as an external knowledge base. Domain adaptation techniques used in NMT usually require fine-tuning and in-domain training data, which may pose difficulties for those working with lesser-resourced languages and may also lead to performance decay of the NMT system for out-of-domain sentences. Scylla does not require fine-tuning of the NMT model, avoiding the risk of model over-fitting and consequent decrease in performance for out-of-domain translations. Two versions of Scylla are presented: one using the source sentence as input, and another one using the target sentence. We evaluate Scylla in comparison to a state-of-the-art commercial NMT system in an experiment in which 50 sentences from the Sports domain are translated from Brazilian Portuguese to English. The two versions of Scylla significantly outperform the baseline commercial system in HTER.