MT4All: Unsupervised MT for Low-resourced language pairs
Description
MT4All aims to create bilingual resources and translation models for language pairs that lack sufficient parallel corpora, by leveraging recent research carried out in the field of unsupervised learning. In particular, MT4All will derive bilingual dictionaries, language models and translation models between English and the following languages: Finnish. Norwegian, Latvian, Basque, Catalan, German, Ukrainian, Georgian, Kazakh and Biomedical Spanish, using large amounts of monolingual corpora only.