Data Migrations

Abstract

Social media platforms offer a rich repository of crowdsourced information that has the potential to monitor human rights violations. The challenge is to quantify, interpret, and situate such unstructured data streams in the broader context, which remains under-investigated in existing CSCW research. Addressing these challenges demands computational solutions to extract large volumes of data in conjunction with human intervention to transition the data streams into the offline context to render them usable and actionable. Following an iterative human-in-the-loop computational approach, we explore whether citizen reports of abductions concentrated on Facebook groups can be useful to complete official records on the ongoing crisis of disappearances in Mexico. We conceptualize three key practices of the process of transitioning the data from online to offline, followed by seven qualitative characteristics of the data streams that contribute to each stage of the process. Our research contributes with an initial understanding of the challenges and opportunities of migrating the local knowledge from online communities to be used as evidence by organizations seeking to address institutional failures.