Training a Bilingual Language Model by Mapping Tokens onto a Shared Character Space
This work addresses the challenge of cross-lingual knowledge for machine translation between Arabic and Hebrew, but it is incremental as it builds on existing language model techniques with a script-specific adaptation.
The authors tackled the problem of training a bilingual Arabic-Hebrew language model by transliterating Arabic into Hebrew script to create a shared character space, resulting in a model that outperforms a baseline using Arabic script and delivers comparable machine translation performance despite using 60% less data.
We train a bilingual Arabic-Hebrew language model using a transliterated version of Arabic texts in Hebrew, to ensure both languages are represented in the same script. Given the morphological, structural similarities, and the extensive number of cognates shared among Arabic and Hebrew, we assess the performance of a language model that employs a unified script for both languages, on machine translation which requires cross-lingual knowledge. The results are promising: our model outperforms a contrasting model which keeps the Arabic texts in the Arabic script, demonstrating the efficacy of the transliteration step. Despite being trained on a dataset approximately 60% smaller than that of other existing language models, our model appears to deliver comparable performance in machine translation across both translation directions.