The fellows (Amelia Taylor Kevin Degila👨🏾‍💻 David Ifeoluwa Adelani 🇳🇬 Thierno Ibrahima DIOP Baamtu @icompass Ghana Natural Language Processing (NLP) The Makerere AI and Data Science Research Lab Ari Ramkilowan) THANK YOU! And my co-conspirator Davor Orlic who was proper in the trenches with me through it all... THANK YOU!

#kebutu Masakhane is an organisation which gives itself the mission to strengthen and spur NLP research by Africans, for Africans. Why not the same kind of organisation for Wolof? masakhane.io

We will be sharing profiles of people actively working Wolof processing. These profiles will also be showcased on our Github repository.

Une question épineuse sur le détournement de l'open source par les Big Tech et l'appropriation du travail des chercheurs africains. Faut il garder pour soit le fruit de son travail (ce qui serait à l'encontre de la recherche) ou l'ouvrir au risque de se le faire "voler" !? 🙆‍♂️🤔

#kebetu Thanks to ÐɛЯǤưƎηE  for restructuring and completing available Wolof text and audio resources github.com/WolofProcessin…

Want to conduct an NLP project on #Wolof ✍️ ? We gather all available datasets in this language as well as possible sources of data in Wolof that could be useful to you 📂 And above all, we are open to contributions 🤝 Thanks to Silèye BA for this great initiative 🙌

VRAIMENT BESOIN DE VOTRE AVIS SVP 🙏 Si vous aviez un Google Assistant (ou Siri ou Amazon Alexa...) capable de comprendre le Wolof (ou votre langue natale), quels genres de tâches aimeriez vous qu'il soit capable de traiter !? 🤔 #kebetu

NLP 101: playing with a corpus of 100 wolof texts. Displaying here the clould of the global word occurence. The bigger, the more its frequent.

Then count the number of document in which a word occurs. The bigger or word is the more documents it occurs

Then finally, displaying the word occurence normalized by their document occurence. This gives less weights to word that occur in many document.