README.md

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)

Web sources:

• Lagos-NWU conversational corpus
• Bíbélì Mímọ́ ní Èdè Yorùbá Òde-Òní
• The Yorùbá blog
• BBC Yorùbá
• Yorùbá for Academic Purpose
• Yobá mọ oduá
• Àwa Ẹlẹ́rìí Jèhófà
• Orí Kìíní
• Iwé ti Nicé
• Alákọ̀wé
• lds.org
• Èdè Yorùbá Rẹwà
• Ìmọ̀_Ẹ̀rọ
• ọ̀rọ̀yorùbá
• Wikipedia
• Asubiaro, T., Adegbola, T., Mercer, R. and Ajiferuke, I. (2018). A Word-Level Language Identification Strategy for Resource-Scarce Languages

Social Media sources:

• https://twitter.com/yobamoodua
• https://twitter.com/yoruba_proverbs
• https://www.facebook.com/oweyoruba

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

• https://clas.uiowa.edu/dwllc/allnet/yoruba-language-and-culture-resources
• https://glosbe.com/yo/en