The study presented in the paper investigates the unintended biases in ASR software against Arabic-speaking refugees, revealing critical insights into how societal religious prejudices are mirrored in technology. The findings of Tuba Bircan and Duha Ceylan highlight the necessity for culturally sensitive approaches in the development of technological tools.
The article is available here.
Full reference: Bircan, T., & Ceylan, D. (2024). Machine Discriminating: Automated Speech Recognition Biases in Refugee Interviews. Journal of Immigrant & Refugee Studies, 1–16. https://doi.org/10.1080/15562948.2024.2355613
Photo: New Africa by Adobe Stock
