Strategic Social Team Crowdsourcing: Forming a Team of Truthful Workers for Crowdsourcing in Social Networks.
Wanyuan Wang,Zhanpeng He,Peng Shi,Weiwei Wu,Yichuan Jiang,Bo An,Zhifeng Hao,Bing Chen
DOI: https://doi.org/10.1109/tmc.2018.2860978
IF: 6.075
2018-01-01
IEEE Transactions on Mobile Computing
Abstract:With the increasing complexity of tasks that are crowdsourced, requesters need to form teams of professional workers that can satisfy complex task skill requirements. Team crowdsourcing in social networks (SNs) provides a promising solution for complex task crowdsourcing, where the requester hires a team of professional workers that are also socially connected can work together collaboratively. Previous social team formation approaches have mainly focused on the algorithmic aspect for social welfare maximization; however, within the traditional objective of maximizing social welfare alone, selfish workers can manipulate the crowdsourcing market by behaving untruthfully. This dishonest behavior discourages other workers from participating and is unprofitable for the requester. To address this strategic social team crowdsourcing problem, truthful mechanisms are developed to guarantee that a worker's utility is optimized when he behaves honestly. This problem is proved to NP-hard, and two efficient mechanisms are proposed to optimize social welfare while reducing time complexity for different scale applications. For small-scale applications where the task requires a small number of skills, a binary tree network is first extracted from the social network, and a dynamic programming-based optimal team is formed in the binary tree. For large-scale applications where the task requires a large number of skills, a team is formed greedily based on the workers' social structure, skill, and working cost. For both mechanisms, the threshold payment rule, which pays each worker his marginal value for task completion, is proposed to elicit truthfulness. Finally, the experimental results of a real-world dataset show that compared to the benchmark exponential VCG truthful mechanism, the proposed small-scale-oriented mechanism can reduce computation time while producing nearly the same social welfare results. Furthermore, compared to other state-of-the-art polynomial heuristics, the proposed large-scale-oriented mechanism can achieve truthfulness while generating better social welfare outcomes.