According to the UN Refugee Agency (UNHCR), Uganda hosted in 2017 over one million refugees. It is Africa’s largest refugee-hosting country and one of the most favourable refugee protection environments in the world, providing displaced people with freedom of movement, the right to work, and access to social services through a generous asylum policy. To the north of Uganda is South Sudan, where years of war, chronic civil instability and famine have forced almost one million South Sudanese to seek refuge across the border and more than two-thirds have arrived in Uganda since the outbreak of violence in capital city Juba started in July 2016.
Pulse Lab Kampala was tasked by the UN in Uganda with understanding the attitudes of host communities towards refugees. The project used the Radio Content Analysis Tool, an automated speech-to-text technology developed by the Lab for less known languages, to analyse public discussions aired by local radio stations to support the Government’s refugee open-door policy. The hypothesis was that the systematic analysis of what people say on the radio regarding their situation, concerns, and needs provides actionable insights for programme implementation.
Results demonstrated that the testimonials, rumours, opinions and reports expressed on the radio by citizens can provide useful insights for early warning systems, monitoring the implementation of projects and programmes, and programme evaluation.
The analysis focused on public opinions expressed in the two main vernacular languages of Uganda: Acholi, spoken in Northern Uganda and Luganda, spoken in the Central Region of the country as well as in Kampala, the capital city. In addition, an early version of automated speech-to-text technology for four other languages namely Rutoro, Lugbara, Ngakaramojong and Runyankore was used to enable the analysis of discussions from West Nile, Rwenzori, Karamoja and the west of Uganda.
The Radio Content Analysis Tool, developed by Pulse Lab Kampala and partners, is an artificial intelligence technology that transcribes audio content into text, which can then be analysed, using machine learning, for topics of interest relevant to the Sustainable Development Goals (SDGs).
The methodology, case studies and findings of the Radio Content Analysis tool are detailed in the report below.
