A Two-staged task assignment algorithm for worker recommendation in a crowdsourcing environment
Rong Chen, Shifei Chen, Xiaoyao Zhang · 2017
Crowdsourcing markets have provided a means to complete tasks cheaply by global netizens. However, having less control of their workers may adversely affect the quality of work, even worse when assigning tasks without considering workers' skills, abilities and commitments. We extend the Dual Task Assigner (DTA) algorithm with the Artificial Bee Colony (ABC) algorithm, which figures out the optimal task baseline level for each task to guide the task assignment to workers with appropriate skills. We empirically evaluate the proposed algorithm using data collected from several crowdsourcing practices by freshmen in computer classes. The results show that our algorithm guarantees task results, and statistically performs better than the existing DTA and random assignment algorithms.