Privacy-Preserved Data Disturbance and Truthfulness Verification for Data Trading
Man Zhang,Xinghua Li,Yinbin Miao,Bin Luo,Wanyun Xu,Yanbing Ren,Robert H. Deng
DOI: https://doi.org/10.1109/tifs.2024.3402162
IF: 7.231
2024-05-25
IEEE Transactions on Information Forensics and Security
Abstract:The advanced data trading allows data generator's (DG) disturbed data to be traded as both initial and reselling trading modes, which meets DG's raw data privacy and data consumers' (DCs) vast data requirement. However, the traded data truthfulness verifiability cannot be guaranteed in the privacy-preserved way. Firstly, due to DG's independent and random disturbance, DC cannot verify whether the traded data is disturbed under his required disturbance parameter without carrying privacy leakage on DG. Secondly, because the reselling trading is allowed, DC can hardly verify the traded data's origin truthfulness under the deceiving of data reseller (DR) while protecting his purchase privacy. Aiming at the above problems, we propose the privacy-preserved data disturbance and truthfulness verification for data trading. Specifically, an honest-but-curious trading server (TS) is introduced to assist our devised private-verifiable imprint-embedded disturbance method where imprint is blinding. Subsequently, TS implements the adaptive truthfulness verification by constructing imprint-embedded individual verification formula and requiring verified participants to decrypt the formula result. The verified participants cannot inform the blinding imprint value to forge the correct result, ensuring the accuracy of the devised verification method. Theoretical analysis proves that participants' privacy is preserved and the traded data's truthfulness can be guaranteed. Extensive experiments using the real-world dataset demonstrate that without any extra privacy cost, our scheme verifies 100% untruthful traded data compared with the existing solutions' 50%.
computer science, theory & methods,engineering, electrical & electronic