Abstract
AbstractElectoral violence is conceived of as violence that occurs contemporaneously with elections, and as violence that would not have occurred in the absence of an election. While measuring the temporal aspect of this phenomenon is straightforward, measuring whether occurrences of violence are truly related to elections is more difficult. Using machine learning, we measure electoral violence across three elections using disaggregated reporting in social media. We demonstrate that our methodology is more than 30 percent more accurate in measuring electoral violence than previously utilized models. Additionally, we show that our measures of electoral violence conform to theoretical expectations of this conflict more so than those that exist in event datasets commonly utilized to measure electoral violence including ACLED, ICEWS, and SCAD. Finally, we demonstrate the validity of our data by developing a qualitative coding ontology.
Highlights
Electoral violence is conceived of as violence that occurs contemporaneously with elections, and as violence that would not have occurred in the absence of an election
We describe the classification accuracy of our neural network compared to a baseline, derive the total number of violent events for each election, and compare the number of events discovered by our neural network to those reported in other event datasets including ACLED, ICEWS, and SCAD
To ensure the neural network has discovered violent events that are correlated with the electoral process, we developed a qualitative coding ontology of all events discovered by our neural network as well as all events recorded by ACLED, ICEWS, and SCAD
Summary
Using textual sources of data to develop estimates of political violence using automated methods is not a new endeavor. Zeitzoff (2011) collected social media data from Twitter to analyze temporal violent dynamics between Israel and Hamas during the 2008–2009 Gaza conflict, and Ramakrishnan et al (2014) used social media data to forecast civil unrest across multiple countries Thanks to these automated methods and massive sources of textual data, scholars of political violence have access to massive datasets measuring political cooperation and conflict. Automated methods code massive amounts of information regarding political violence, including outbreaks of civil and international conflict (D’Orazio et al, 2014), and have identified perpetrators of mass atrocities (Bagozzi and Koren, 2017) These algorithms, including neural networks, have achieved accuracy beyond that of previously utilized textual analysis methods, such as parsers (Beieler, 2016; Lin et al, 2016). With more accurate discovery of these events, better statistical models can be constructed to inform scholars of the mechanisms underlying such violence and its impacts on society
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
