Publication

Abstract : Crowdsourcing is an emerging technology which enables human workers to perform the task that cannot be accomplished by automated tools. The research in crowdsourcing mainly focuses on quality control, workflow optimization, privacy preserving and trust management. However, one of the greatest challenges for the researchers is the unavailability of real dataset from crowdsourcing platforms suitable for research. This paper analyzes the characteristics of workload generated in crowdsourcing systems using trace-driven approach. We study the behaviour of tasks and workers arriving at the crowdsourcing platforms using spatial and temporal features. For this study both emprical and synthetic workload are used. Moreover, for reproducible research, we introduce a workload generator for crowdsourcing platforms which generates an unbiased workload and mimics the real crowdsourcing system workload. In addition, the tool features the statistical analysis of the crowdsourcing dataset and it is suitable for studies such as quality control and productivity assessment.