Predicting Hate Speech for Hausa African Indigenous Language

The region of West Africa still lacks more research on hate speech in their regional languages, while having a historical problem with ethnic conflicts, many times related to racism and terrorism. One example consists of Nigeria country. During independence, the British divided Nigeria primarily along ethnic lines. The north with a Hausa and Fulani majority, Yoruba being concentrated in the southwest, and the Igbo in the southeast. Take into consideration this division, in the southeastern part of the country, there was a secessionist unrecognized state from 1967 to 1970, named the Republic of Biafra. Its existence has an association with Igbo nationalism, even though other ethnic peoples of the area, which more recently sparked renewed conflicts in the region for the independence of Biafra. Besides Boko Haram in the north, a set of groups from his region were considered terrorists by the Nigerian government. To fill this important research gap and mitigate this indisputably relevant social problem, in this project, we focus on the investigation and development of resources and methods for hate speech detection in the Hausa African Indigenous Language.


Team
  • Francielle Vargas. Institute of Mathematics and Computer Sciences, University of São Paulo, Brazil
  • Samuel Guimarães. Computer Science Departament, Federal University of Minas Gerais, Brazil
  • Diego Alves . Department of Language and Technology, Saarland University, Germany
  • Shamsuddeen Hassan . Computer Science Departament, University of Lisboa, Portugal
  • Diallo Mohamed. Institute of Applied Social Sciences, University of St. Thomas Aquinas, Burkina Faso
  • Ibrahim Said Ahmad. Department of Information Technology, Bayero University Kano, Nigeria
  • Idris Abdulmumin. Department of Information Technology, Bayero University Kano, Nigeria
  • Fabrício Benevenuto. Computer Science Departament, Federal University of Minas Gerais, Brazil

Publications

Resources
Dataset

Sponsorship