Niger-Volta-LTI/yoruba-text

Yorùbá language training text for NLP, ASR and TTS tasks # Yorùbá text This repository contains fully diacritized Yorùbá text, converted to Unicode Normalization Form Composition (NFC) format, where diacritized characters are _composed_ into a single character with the following code: ``` def convert_to_NFC(filename, outfilename): text=''.join(c for c in unicodedata.normalize('NFC', open(filename).read())) with open(outfilename, 'w') as f: f.write(text) ``` ### Sources: * Lagos-NWU conversational corpus * Bíbélì Mímọ́ ní Èdè Yorùbá Òde-Òní * The Yorùbá blog * Asubiaro, T., Adegbola, T. et al. (2018). A Word-Level Language Identification Strategy for Resource-Scarce Languages * Òwe Yorùbá * Ìwé Ti Mọ́mọ́nì * Kùránì (Qur'an) Mímọ́ #### Sources yet to be scraped and cleaned * BBC Yorùbá * Yorùbá for Academic Purpose * Yobá mọ oduá * Àwa Ẹlẹ́rìí Jèhófà * Orí Kìíní * Iwé ti Nicé * Alákọ̀wé * Èdè Yorùbá Rẹwà * Ìmọ̀_Ẹ̀rọ * ọ̀rọ̀yorùbá * Wikipedia * Poetry of Ọláńrewájú Adépọ̀jù #### Social Media sources: * twitter.com * twitter.com * facebook.com Text has been gathered with permission from online sources, and lightly preprocessed for use in NLP, TTS, ASR applications. Note, some of the sentences may have errors, please submit a pull-request if you have corrections! ## Resources * clas.uiowa.edu * glosbe.com ## Bibtex If you want to cite this repo in your work, please use: ``` @misc{Orife_yoruba-text_2018, author = {Orife, Iroro and Fasubaa, Timilehin and Wahab, Olamilekan}, month = {1}, title = {{yoruba-text}}, url = {github.com, year = {2018} } ```