Investigating Lexical Sharing in Multilingual Machine Translation for Indian Languages
This work addresses translation efficiency for Indian languages, but it is incremental as it builds on existing multilingual model strategies without major breakthroughs.
The study investigated lexical sharing in multilingual machine translation from Hindi, Gujarati, and Nepali into English, finding that transliteration did not significantly improve performance and that models trained on original scripts were robust to cross-script differences, even for low-resource languages like Marathi and Bengali.
Multilingual language models have shown impressive cross-lingual transfer ability across a diverse set of languages and tasks. To improve the cross-lingual ability of these models, some strategies include transliteration and finer-grained segmentation into characters as opposed to subwords. In this work, we investigate lexical sharing in multilingual machine translation (MT) from Hindi, Gujarati, Nepali into English. We explore the trade-offs that exist in translation performance between data sampling and vocabulary size, and we explore whether transliteration is useful in encouraging cross-script generalisation. We also verify how the different settings generalise to unseen languages (Marathi and Bengali). We find that transliteration does not give pronounced improvements and our analysis suggests that our multilingual MT models trained on original scripts seem to already be robust to cross-script differences even for relatively low-resource languages