Papers with rare word problem
Multimodal Machine Translation with Embedding Prediction (N19-3)
Copied to clipboard
| Challenge: | Pretrained word embeddings improve multimodal machine translation of low-resource domains due to a shortage of training data. |
| Approach: | They propose to combine pretrained word embeddings with search-based approaches to improve NMT of low-resource domains to better translate rare words. |
| Outcome: | The proposed approach improves translation performance by 1.24 METEOR and 2.49 BLEU and achieves 7.67 F-score. |
Handling Rare Word Problem using Synthetic Training Data for Sinhala and Tamil Neural Machine Translation (L18-1)
Copied to clipboard
| Challenge: | Lack of parallel training data influences rare word problem in Neural Machine Translation systems, especially for underresourced languages. |
| Approach: | They propose to use Parts of Speech tagging and morphological analysis as syntactic features to prune generated synthetic sentence pairs that do not adhere to language syntax. |
| Outcome: | The proposed methods show that they can prune sentences that do not adhere to language syntax over Sinhala to Tamil and Tamil to Sinhalak translation systems. |