IoT Privacy-Preserving Data Mining With Dynamic Incentive Mechanism

With the rise of the Internet of Things (IoT), IoT data analytics has gradually stepping into the spotlight of data mining. Existing research has primarily focused on enhancing the precision of IoT data mining, while the privacy protection aspects have not been fulfilled so far. The deployment of IoT data mining is contingent on the protection of data privacy and its economic worth to all parties. However, most existing IoT data mining research disregards economic benefits and lacks incentives, limiting its applicability. To address this issue, we provide a system for differential privacy-based IoT privacy-preserving data mining (IoT-PPDM) with dynamic incentive mechanism, and a formal economic model for IoT data mining is constructed. We utilized noncooperative game theory to simulate the multilateral interaction process in IoT data mining. To encourage participation from all parties, a dynamic incentive mechanism is designed to establish a balance between privacy protection and data mining requirements. In addition, we discuss the utility of all participants and theoretically validate the feasibility of IoT-PPDM. The experimental results show that IoT-PPDM with dynamic incentive mechanism can increase the benefits for all participants while avoiding irrational behavior of all parties.