KatyaAplonova/UD_Bambara

# Summary The UD Bambara treebank is a section of the Corpus Référence du Bambara annotated natively with Universal Dependencies. # Introduction Bambara (also known as Bamana) is the most widely-spoken language of the Manding language group (Niger-Congo > Mande > Western Mande). It is spoken mainly in Mali by 13-14 million people; of these, around four million are L1 speakers. Development of the Bambara Reference Corpus (cormand.huma-num.fr) was started in April 2012 [Vydrin 2013, Maslinsky 2014]. The corpus includes a non-disambiguated sub-corpus and a disambiguated one. For a current moment, the whole corpus contains about 9 million tokens. The corpus was annotated using UD Annotatrix annotation tool [Tayers, Sheyanova, Washington 2018]. Documentation for a treebank is available on UD site (universaldependencies.org). ## References Maslinsky, K. (2014). Daba: a model and tools for Manding corpora. In Proceedings of TALAf 2014 : Traitement Automatique des Langues Africaines, pages 114-122. Tyers, F. M., Sheyanova, M., and Washington, J. N. (2018). UD Annotatrix: An annotation tool for Universal Dependencies. In Proceedings of the 16th International Workshop on Treebanks and Linguistic Theories. Vydrin, V. (2013). Bamana reference corpus (BRC). Procedia - Social and Behavioral Sciences, 95, pages 75–80. # Source files for UD_Bambara * `html/`: the original `.html` files from the CBR * `scripts/`: helper scripts for converting the data from CBR to UD (morphology and POS) `.conllu` files in this directory are the result of manual annotation of the output of the conversion script. # Acknowledgments The conversion and annotation has been done by Katya Aplonova and Francis M. Tyers at the Higher School of Economics in Moscow. We would like to thank the developers and annotators of the Corpus Référence du Bambara for permission to base this on their work. === Machine-readable metadata (DO NOT REMOVE!) ================================ Da …