Responsible AI, bias and fairness

A half-day introduction to responsible AI and bias in machine translation models for African languages, in particular. There is significant growth in the area of building Natural Language Processing (NLP) tools for African languages, despite the lower availability of explicit text data than for English. In this masterclass we introduce concepts on Responsible AI and then dive into building models for machine translation models for African languages. We then discuss the approach taken to evaluate and quantify the gender bias within a Luganda-English machine translation system using Word Embeddings Fairness Evaluation Framework (WEFE).

Teachers: Dr. Joyce Nakatumba-Nabende (Makerere University) and Dr. Andrew Katumba (Makerere University)