Spatial Crowdsourcing Quality Control Model Based on K-Anonymity Location Privacy Protection and ELM Spammer Detection

The spatial crowdsourcing task places workers at a risk of privacy leakage. If positional information is not required to submit, it will result in an increased error rate and number of spammers, which together affects the quality of spatial crowdsourcing. In this paper, a spatial crowdsourcing quality control model is proposed, called SCQCM. In the model, the spatial k-anonymity algorithm is used to protect the position privacy of the general spatial crowdsourcing workers. Next, an ELM (extreme learning machine) algorithm is used to detect spammers, while an EM (expectation maximization) algorithm is used to estimate the error rate. Finally, different parameters are selected, and the efficiency of the model is simulated. The results showed that the spatial crowdsourcing model proposed in this paper guaranteed the quality of crowdsourcing projects on the premise of protecting the privacy of workers.

[1]  Panos Kalnis,et al.  SABRE: a Sensitive Attribute Bucketization and REdistribution framework for t-closeness , 2011, The VLDB Journal.

[2]  Chee Kheong Siew,et al.  Extreme learning machine: Theory and applications , 2006, Neurocomputing.

[3]  José A. Pino,et al.  Computer supported collaborative work j.Ucs special issue , 2016 .

[4]  Fuzhen Zhuang,et al.  Clustering in extreme learning machine feature space , 2014, Neurocomputing.

[5]  Chi-Yin Chow,et al.  Spatial cloaking for anonymous location-based services in mobile peer-to-peer environments , 2011, GeoInformatica.

[6]  Latanya Sweeney,et al.  k-Anonymity: A Model for Protecting Privacy , 2002, Int. J. Uncertain. Fuzziness Knowl. Based Syst..

[7]  Cyrus Shahabi,et al.  A Framework for Protecting Worker Location Privacy in Spatial Crowdsourcing , 2014, Proc. VLDB Endow..

[8]  Qi Han,et al.  Spatial crowdsourcing: current state and future directions , 2016, IEEE Communications Magazine.

[9]  Hisashi Kashima,et al.  Preserving worker privacy in crowdsourcing , 2014, Data Mining and Knowledge Discovery.

[10]  Ying-Jie Wu,et al.  Algorithm for k -Anonymity Based on Rounded Partition Function: Algorithm for k -Anonymity Based on Rounded Partition Function , 2012 .

[11]  Guang-Bin Huang,et al.  An Insight into Extreme Learning Machines: Random Neurons, Random Features and Kernels , 2014, Cognitive Computation.

[12]  A. P. Dawid,et al.  Maximum Likelihood Estimation of Observer Error‐Rates Using the EM Algorithm , 1979 .

[13]  Devavrat Shah,et al.  Budget-Optimal Task Allocation for Reliable Crowdsourcing Systems , 2011, Oper. Res..

[14]  D. Rubin,et al.  Maximum likelihood from incomplete data via the EM - algorithm plus discussions on the paper , 1977 .

[15]  Zhipeng Cai,et al.  A differentially k-anonymity-based location privacy-preserving for mobile crowdsourcing systems , 2017, IIKI.

[16]  Cheng Wu,et al.  Semi-Supervised and Unsupervised Extreme Learning Machines , 2014, IEEE Transactions on Cybernetics.