Received: 01 March 2016 / Accepted: 07 September 2016/ Published online: 07 November 2016
Digital volunteers were found to be decentralized, diverse, and independent.
The four “types of crowdsourced crisis-related information” were: 1) community-based, 2) infrastructure-based, 3) information-based, and 4) future threats. These were shared through five “types of tasks”: 1) text/image data submission; 2) text/image sorting; 3) text/image mapping; 4) object tagging; and 5) donation/ crowdfunding.
Digital humanitarian outputs evolved in three phases: 1) first degree material outputs; 2) non-material outputs; and 3) second degree material output.
Humanitarian websites presented two crowdsourcing archetypes: 1) crowd creation to crowd solving; and 2) crowd rating to crowd processing.
A humanitarian crowdsourcing model is proposed.
Lately, the humanitarian community has been utilizing crowdsourcing to facilitate medical and disaster response. Grounded in Geiger et al.’s (2011) Crowdsourcing Information Systems (CIS) and Suroweicki’s (2004) Wisdom of the Crowds (WC), this study content-analyzed 23 humanitarian crowdsourcing websites to find out how crowdsourcing has enabled medical and disaster response, as evident in global humanitarian movements from 2010 to 2014. Findings revealed that the digital volunteers that generate big data in humanitarian crowdsourcing websites are composed of an independent crowd with diverse opinions and a decentralized demography. The crisis-related inputs they contributed were community-based, infrastructure-based, information-based, and related to future threats. Emerging outputs resulted in three phases: 1) first degree material outputs (e.g., geographic crisis map, and/or text database/resource page); 2) non-material outputs (e.g., planning, strategizing, and operationalization via partner organizations); and 3) second degree material output (e.g., actual medical and disaster response). Results of the analysis suggest that the dynamic process of humanitarian crowdsourcing do not appear to be boxed into the CIS archetypes. Instead, the characteristic of the websites presented a merging of the four archetypes into two: 1) crowd creation and crowd solving; and 2) crowd rating and crowd processing. As an outcome of the study, a humanitarian crowdsourcing model is proposed.