Open source data reveals connection between online and on-street protest activity

Hong Qi, Pedro Manrique, Daniela Johnson, Elvira Restrepo, Neil F. Johnson

Research output: Contribution to journalArticle

7 Scopus citations

Abstract

There is enormous interest in inferring features of human behavior in the real world from potential digital footprints created online - particularly at the collective level, where the sheer volume of online activity may indicate some changing mood within the population regarding a particular topic. Civil unrest is a prime example, involving the spontaneous appearance of large crowds of otherwise unrelated people on the street on a certain day. While indicators of brewing protests might be gleaned from individual online communications or account content (e.g. Twitter, Facebook) societal concerns regarding privacy can make such probing a politically delicate issue. Here we show that instead, a simple low-level indicator of civil unrest can be obtained from online data at the aggregate level through Google Trends or similar tools. Our study covers countries across Latin America during 2011-2014 in which diverse civil unrest events took place. In each case, we find that the combination of the volume and momentum of searches from Google Trends surrounding pairs of simple keywords, tailored for the specific cultural setting, provide good indicators of periods of civil unrest. This proof-of-concept study motivates the search for more geographically specific indicators based on geo-located searches at the urban level.

Original languageEnglish (US)
Article number18
JournalEPJ Data Science
Volume5
Issue number1
DOIs
StatePublished - Dec 1 2016

Keywords

  • civil unrest
  • online searches
  • open source data
  • social media

ASJC Scopus subject areas

  • Modeling and Simulation
  • Computer Science Applications
  • Computational Mathematics

Fingerprint Dive into the research topics of 'Open source data reveals connection between online and on-street protest activity'. Together they form a unique fingerprint.

  • Cite this