Mon. Mar 4th, 2024

TLDR:

– Ife Adebara, a programmer and scholar at the University of British Columbia, is working to make AI technology more accessible for speakers of over 500 African languages.
– Adebara’s project, called Afrocentric Natural Language Processing, aims to create awareness, tools, and programs that are accessible to speakers of African languages, including Swahili and Zulu.

Scientist Ife Adebara is on a mission to make AI technology accessible to all people in Africa by integrating their indigenous languages with the technology they use. Adebara, a programmer and scholar at the University of British Columbia, is leading a project called Afrocentric Natural Language Processing, which works to create awareness, tools, and programs that are accessible to speakers of African languages.

Adebara notes that AI technology is rapidly evolving but often leaving behind speakers of languages other than English. She explains that individuals who speak minority languages often have to put their language aside to access technology in English, leading to a decline in the use of their native language and language endangerment. The Afrocentric Natural Language Processing project aims to mitigate this issue.

Currently, Adebara’s team has released two language identification programs online called SERENGETI and AfroLID, which help identify African languages. Adebara’s project focuses on approximately 517 languages spoken in 50 out of 54 African countries, with the goal of expanding to as many languages as possible on the continent.

One challenge Adebara faces is that many African languages are considered low-resource languages, meaning there is insufficient data to build classical language models for AI. Adebara addresses this challenge by developing models that learn from multiple languages, increasing the dataset and improving performance. However, she acknowledges that more data is needed to achieve near-human accuracy.

Adebara emphasizes the importance of including African languages in the development of AI technology. Excluding African languages means excluding more than one billion people in Africa from global conversations in their indigenous languages. Additionally, African languages have unique grammatical features that aren’t learned by models and technologies that exclude them. Adebara believes that making technology accessible in African languages will have long-term effects on education and provide access to information, health resources, and tools like Google Maps.

In summary, Ife Adebara is pioneering efforts to integrate AI technology with over 500 African languages. Through her project Afrocentric Natural Language Processing, Adebara aims to make technology accessible to all African people by developing assets, building AI models, and deploying them in indigenous languages. By including African languages, Adebara hopes to bridge the digital divide and ensure that AI technology benefits speakers of diverse languages.