MHDP-Project/mental-health-nlp-uganda-tanzania

Speech recognition and NLP tools for mental health telephone services in East Africa, supporting English, Luganda, and Swahili # NLP for Mental Health Call Centers in Uganda and Tanzania. Developing speech recognition and NLP tools for mental health telephone services in East Africa, supporting English, Luganda, and Swahili. ## About This project develops automatic speech recognition (ASR) and natural language processing (NLP) models and datasets for mental health services in East Africa. We use call conversation recordings from hospital call centers as primary data sources. The project aims to provide insights into the role of communication quality in mental healthcare and the feasibility of using digital health and AI technologies to improve mental health services. The datasets and models developed in this project will also facilitate the advancement of NLP for under-resourced African languages. The project focuses on Luganda, Swahili, and English. ## Related Publications - Kabukye JK, et al. (2024). Implementing a Hospital Call Center Service for Mental Health in Uganda: User-Centered Design Approach. *JMIR Human Factors*. Link - Nakatumba-Nabende J, et al. (2024). Building Text and Speech Benchmark Datasets and Models for Low-Resourced East African Languages. *Applied AI Letters*. Link