TY - GEN
T1 - Crowdsourcing in the Field
T2 - 3rd AAAI Conference on Human Computation and Crowdsourcing, HCOMP 2015
AU - Agapie, Elena
AU - Teevan, Jaime
AU - Monroy-Hernández, Andrés
N1 - Funding Information:
We thank Microsoft Research FUSE Labs for funding, Melissa Quintanilha - design, Todd Newman - development, Daniel Epstein, Neha Kumar, MSR nexus group - feedback, Katya Yefimova, Gary Hsieh, Khai Troung - deployments.
Publisher Copyright:
Copyright © 2015, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved.
PY - 2015/11/12
Y1 - 2015/11/12
N2 - While crowd work typically involves tasks that performed at any time and anywhere, some tasks inherently require the physical presence of workers at a specific time and location. This paper presents a case study of a hybrid crowdsourcing process that involves the collaborative production of event reports using a combination of local and remote workers. The process extends human computation into the physical world by using local workers to collect information in person at events and remote workers to curate the collected information and generate event reports. We deployed the process at 11 events, employing 84 workers, and identified the challenges local workers face as constraints in mobility, time available to perform tasks, unpredictability of events, and interaction with others. We discuss issues related to collaboration with remote workers and bias in field reporting, and conduct a qualitative analysis to make design recommendations for extending human computation into the physical environment.
AB - While crowd work typically involves tasks that performed at any time and anywhere, some tasks inherently require the physical presence of workers at a specific time and location. This paper presents a case study of a hybrid crowdsourcing process that involves the collaborative production of event reports using a combination of local and remote workers. The process extends human computation into the physical world by using local workers to collect information in person at events and remote workers to curate the collected information and generate event reports. We deployed the process at 11 events, employing 84 workers, and identified the challenges local workers face as constraints in mobility, time available to perform tasks, unpredictability of events, and interaction with others. We discuss issues related to collaboration with remote workers and bias in field reporting, and conduct a qualitative analysis to make design recommendations for extending human computation into the physical environment.
UR - http://www.scopus.com/inward/record.url?scp=85015774101&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85015774101&partnerID=8YFLogxK
U2 - 10.1609/hcomp.v3i1.13235
DO - 10.1609/hcomp.v3i1.13235
M3 - Conference contribution
AN - SCOPUS:85015774101
T3 - Proceedings of the 3rd AAAI Conference on Human Computation and Crowdsourcing, HCOMP 2015
SP - 2
EP - 11
BT - Proceedings of the 3rd AAAI Conference on Human Computation and Crowdsourcing, HCOMP 2015
A2 - Gerber, Elizabeth
A2 - Ipeirotis, Panos
PB - AAAI press
Y2 - 8 November 2015 through 11 November 2015
ER -