III: Small: GeoCrowd - A Generic Framework for Trustworthy Spatial Crowdsourcing

University Of Southern California, Los Angeles CA

Investigators

Abstract

Many studies foresee significant growth in the number of smart phone users, the phone's hardware and software features, and the broadband bandwidth. Therefore, a transformative area of research is to utilize this new platform for various tasks, among which the most promising is spatial crowdsourcing. Spatial crowdsourcing engages individuals, groups, and communities in the act of collecting, analyzing, and disseminating urban, social, and other spatiotemporal information. Two major impediments to the success of spatial crowdsourcing in real-world applications are scalability and trust issues. Therefore, the first objective of this project is to study the issue of scale in spatial crowdsourcing. In particular, given that the task assignment is the main bottleneck of the system, the spatial aspects of the tasks are exploited to reduce the complexity of assignment. In addition, a cloud-based distributed approach to the problem is investigated for better scale-out. The second objective is to extend the framework to incorporate trust by maintaining a reputation score per worker and a confidence level for every spatial task. Consequently, multiple workers can perform a task redundantly in order to satisfy its confidence level. The spatial task assignment solutions are extended to take redundant task assignments and confidence satisfaction into consideration. Spatial crowdsourcing has applications in disaster-response, urban planning, intelligent transportation, journalism and intelligence. As part of this project a spatial crowdsourcing system is being developed that can be used for real-world data collection and evaluation as well as for social studies and educational purposes. Results are disseminated through the web: http://infolab.usc.edu/projects/GeoCrowd/