Currently, crowdsourced query processing is done on reward-driven platforms such as Amazon Mechanical Turk (AMT) and Crowd-Flower. However, due to budget constraints for conducting a crowdsourcing task in practice, the scalability is inherently poor. In this paper, we exploit microblogs for supporting crowdsourced query processing. We leverage the social computation power and decentralize the evaluation of the crowdsourcing platforms queries towards social networks. We propose a new problem of minimizing the cost of processing crowdsourced queries on microblogs, given a specified accuracy threshold of users' votes. This problem is NP-hard and its computation is #P-hard. To tackle this problem, we develop a greedy algorithm with a quality guarantee. We demonstrate the performance on real datasets.
展开▼
