Research on Cost Control of Mobile Crowdsourcing Supporting Low Budget in Large Scale Environmental Information Monitoring

Due to the large scale of information, high cost of deployment and vulnerability of instrument, it is difficult to solve the problem of information monitoring in urban environment. The emergence of mobile crowdsourcing solves the problems of high deployment cost and fragile instrument, which makes it possible to solve this monitoring problem. However, the massive employment cost because of the huge information scale makes it difficult for mobile crowdsourcing technology to be applied in practice. This paper proposes a low-budget model based on compressed sensing and naive Bayes classifier, which takes into account the impact of human activities on environmental information, improves the recovery algorithm of compressed sensing algorithm, and reduces the error of data recovery. At the same time, this model also considers the fact that some participants are not qualified to complete the task and improves the naive Bayes classifier to identify more reliable participants to reduce the reemployment rate. The model in this paper can recover all the data with a small number of sampling points, thus greatly reducing the task cost. At the same time, it can identify qualified participants who can complete the crowdsourcing task with high accuracy, thus further reducing the task cost. Experiments show that the low budget model proposed in this paper has a good performance on cost control.