TY - GEN
T1 - Towards Reliable Social Sensing in Cyber-physical-social Systems
AU - Huang, Chao
AU - Marshall, Jermaine
AU - Wang, Dong
AU - Dong, Mianxiong
N1 - This material is based upon work supported by the National Science Foundation under Grant No. IIS-1447795.
PY - 2016/7/18
Y1 - 2016/7/18
N2 - Social sensing is a new application paradigm of cyber-physical-social systems (CPSS), where a group of individuals volunteer to report their claims about the physical environment using cyber devices. A fundamental problem in social sensing application is to ascertain source reliability and the claim correctness without knowing either of them a priori, which is referred to as truth finding. Several key challenges exist in order to solve the truth finding problem. First, neither the source reliability nor the correctness of collected data are known a priori. Second, data sources may make their claims with different degrees of uncertainty and confidence. Third, it is challenging to accurately quantify the quality of truth finding results without knowing the ground truth information. In this paper, we develop a confidence-aware truth finding scheme to address the above challenges under a unified analytical framework. The confidence-aware truth finding scheme solves a constraint estimation problem to jointly estimate both the source reliability and claim correctness by explicitly considering source confidence and uncertainty on the reported claims. To rigorously quantify the accuracy of the MLE estimation, we also derive confidence bounds on the estimation results. Finally, we evaluate our confidence-aware scheme with techniques from current literature through an extensive simulation study. The evaluation results validates the performance gains achieved by our proposed solution compared to other baselines.
AB - Social sensing is a new application paradigm of cyber-physical-social systems (CPSS), where a group of individuals volunteer to report their claims about the physical environment using cyber devices. A fundamental problem in social sensing application is to ascertain source reliability and the claim correctness without knowing either of them a priori, which is referred to as truth finding. Several key challenges exist in order to solve the truth finding problem. First, neither the source reliability nor the correctness of collected data are known a priori. Second, data sources may make their claims with different degrees of uncertainty and confidence. Third, it is challenging to accurately quantify the quality of truth finding results without knowing the ground truth information. In this paper, we develop a confidence-aware truth finding scheme to address the above challenges under a unified analytical framework. The confidence-aware truth finding scheme solves a constraint estimation problem to jointly estimate both the source reliability and claim correctness by explicitly considering source confidence and uncertainty on the reported claims. To rigorously quantify the accuracy of the MLE estimation, we also derive confidence bounds on the estimation results. Finally, we evaluate our confidence-aware scheme with techniques from current literature through an extensive simulation study. The evaluation results validates the performance gains achieved by our proposed solution compared to other baselines.
KW - Confidence-Aware
KW - Cyber-Physical-Social Systems
KW - Maximum Likelihood Estimation
KW - Performance Bounds
KW - Social Sensing
KW - Truth Finding
UR - https://www.scopus.com/pages/publications/84991629013
UR - https://www.scopus.com/pages/publications/84991629013#tab=citedBy
U2 - 10.1109/IPDPSW.2016.132
DO - 10.1109/IPDPSW.2016.132
M3 - Conference contribution
AN - SCOPUS:84991629013
T3 - Proceedings - 2016 IEEE 30th International Parallel and Distributed Processing Symposium, IPDPS 2016
SP - 1796
EP - 1802
BT - Proceedings - 2016 IEEE 30th International Parallel and Distributed Processing Symposium, IPDPS 2016
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 30th IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2016
Y2 - 23 May 2016 through 27 May 2016
ER -