African & Indigenous Futures: Harnessing AI for Global South Languages

This Towards a National Collection (TaNC) will focus on the integration of Indigenous and African languages into computational processes, bringing together Robert Pugh from Indiana University, Bloomington and Ignatius Ezeani from Lancaster University:

• Robert Pugh, Indiana University, Bloomington: Indigenous voices past, present and future: Digital technologies for Nahuatl and Maya
• Ignatius Ezeani, Lancaster University: NLP and Large Language Models: A Case for Inclusive Language Technologies.

Showcasing the role of innovation in preserving and revitalising African and indigenous cultural heritage, these speakers will discuss the dynamic intersection of Human Language Technology (HLT) and inclusive Natural Language Processing (NLP) models. The webinar will address important topics when working within African and indigenous spaces, such as:

• the importance of collaborative efforts
• accessibility of resources
• careful design of bias-aware language models.

Full abstracts

Robert Pugh: Indigenous voices past, present and future: Digital technologies for Nahuatl and Maya

Nahuatl and Maya are the two most widely spoken indigenous languages in Mexico. Along with many other Mesoamerican languages, they share notable sociolinguistic features: wide dialectal and orthographic variation, a rich pre-/post-colonial literary tradition, and extensive language contact with Spanish. In this talk, I discuss the development of Human Language Technology (HLT) for these two languages, highlighting the diverse functions and utilities of HLT in this context, native speaker training and collaboration, and the importance of public/permissively-licensed linguistic resources.

Ignatius Ezeani: NLP and Large Language Models: A Case for Inclusive Language Technologies

Large language models have taken the world by storm recently but how inclusive are they? In today's diverse world, the need for inclusive language technologies cannot be over-emphasized. By acknowledging and addressing biases, promoting representation, and accommodating diverse linguistic and cultural contexts, inclusive language technologies can foster equitable communication and empower marginalized communities. Through awareness, deliberate design choices, and ongoing refinement, we advocate for the development and implementation of inclusive language technologies to promote social justice and enhance communication for all.

Click here for information on Masakhane: a grassroots organisation whose mission is to strengthen and spur NLP research in African languages, for Africans, by Africans.

For more details about Ignatius Ezeani's work see ORCiD or GScholar.