Pronoun-Targeted Fine-tuning for NMT with Hybrid Losses
This addresses the specific issue of pronoun translation accuracy for machine translation systems, but it is incremental as it builds on existing fine-tuning methods.
The paper tackles the problem of improving pronoun translations in neural machine translation without using additional data by introducing a class of conditional generative-discriminative hybrid losses for fine-tuning, resulting in BLEU score improvements such as from 31.81 to 32.00 on WMT14 De-En and from 32.10 to 33.13 on IWSLT13 De-En.
Popular Neural Machine Translation model training uses strategies like backtranslation to improve BLEU scores, requiring large amounts of additional data and training. We introduce a class of conditional generative-discriminative hybrid losses that we use to fine-tune a trained machine translation model. Through a combination of targeted fine-tuning objectives and intuitive re-use of the training data the model has failed to adequately learn from, we improve the model performance of both a sentence-level and a contextual model without using any additional data. We target the improvement of pronoun translations through our fine-tuning and evaluate our models on a pronoun benchmark testset. Our sentence-level model shows a 0.5 BLEU improvement on both the WMT14 and the IWSLT13 De-En testsets, while our contextual model achieves the best results, improving from 31.81 to 32 BLEU on WMT14 De-En testset, and from 32.10 to 33.13 on the IWSLT13 De-En testset, with corresponding improvements in pronoun translation. We further show the generalizability of our method by reproducing the improvements on two additional language pairs, Fr-En and Cs-En. Code available at <https://github.com/ntunlp/pronoun-finetuning>.