Solving team making problem for crowdsourcing with hybrid metaheuristic algorithm
Han Wang, Zhilei Ren, Xiaochen Li, Xin Chen, He Jiang · Proceedings of the Genetic and Evolutionary Computation Conference Companion · 2018
For a typical crowdsourcing process, a task publisher first publishes a task with an acceptable budget. Then hundreds of crowdsourced workers apply for the task with their desired bids. To recruit an adequate Crowdsourced Virtual Team (CVT) while balancing the profits of the task publisher and crowdsourced workers, previous studies proposed various algorithms, including Genetic Algorithm (GA), Alternating Variable Method (AVM), etc. However, the performance is still limited. In this study, we propose a novel hybrid metaheuristic algorithm CVTMaker to help publishers identify ideal CVTs. CVTMaker is effective which combines (1+1) Evolutionary Strategy ((1+1)-ES) and AVM to search solutions. Experimental results show that CVTMaker significantly outperforms GA and AVM over 3,117 and 5,642 of the 6,000 instances respectively.