Structure of ethnic violence in Sudan: a semi-automated network analysis of online news (2003---2010)

  • Authors:

  • Tracy Holt;Jeffrey C. Johnson;James D. Brinkley;Kathleen M. Carley;Janna Caspersen

  • Affiliations:

  • Institute for Coastal Science and Policy, East Carolina University, Greenville, USA 27858 and Geography Department, East Carolina University, Greenville, USA 27858;Institute for Coastal Science and Policy, East Carolina University, Greenville, USA 27858 and Sociology Department, East Carolina University, Greenville, USA 27858;Institute for Coastal Science and Policy, East Carolina University, Greenville, USA 27858;Institute for Software Research, School of Computer Science, Carnegie Mellon University, Pittsburgh, USA 15213;Institute for Coastal Science and Policy, East Carolina University, Greenville, USA 27858 and Geography Department, East Carolina University, Greenville, USA 27858

  • Venue:
  • Computational & Mathematical Organization Theory
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

Mining textual sources of data can be used to design studies and test theories at temporal and spatial scales unheard of in the past. This opens up new opportunities for conflict studies and ethnographic research. We conducted a semi-automated network analysis of the 2003---2010 Sudan Tribune online news articles and modeled ethnic-group conflict in Sudan. We tested whether an ethnic group's connections to the environment (livestock, biomes, and other resources) and other ethnic groups was associated with severe conflict and peace terms and whether ethnic-group richness at a given geospatial location was associated with severe conflict. Ethnic groups with more connections to livestock, environmental resources, and those ethnic groups with denser environmental---resource networks in their sphere of influence had more co-occurrences with severe conflict terms. Ethnic groups that had more connections to biomes in their sphere of influence were co-associated with peace terms. Locations that had high ethnic-group richness were associated with a higher frequency of severe conflict terms.