Cross-lingual Character-Level Neural Morphological Tagging

Ryan Cotterell, Georg Heigold
2017 Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing  
Even for common NLP tasks, sufficient supervision is not available in many languages-morphological tagging is no exception. In the work presented here, we explore a transfer learning scheme, whereby we train character-level recurrent neural taggers to predict morphological taggings for high-resource languages and low-resource languages together. Learning joint character representations among multiple related languages successfully enables knowledge transfer from the high-resource languages to
more » ... urce languages to the low-resource ones, improving accuracy by up to 30%.
doi:10.18653/v1/d17-1078 dblp:conf/emnlp/CotterellH17 fatcat:mlwjbvefsrckvb2eihoppwae6y