View in EDS

Hasse sensitivity level: A sensitivity-aware trajectory privacy-enhanced framework with Reinforcement Learning.

Saved in:
Bibliographic Details
Title: Hasse sensitivity level: A sensitivity-aware trajectory privacy-enhanced framework with Reinforcement Learning.
Authors: Zhang, Jing1 (AUTHOR) jing165455@126.com, Huang, Yi-rui1 (AUTHOR), Huang, Qi-han1 (AUTHOR), Li, Yan-zi1 (AUTHOR), Ye, Xiu-cai2 (AUTHOR)
Source: Future Generation Computer Systems. May2023, Vol. 142, p301-313. 13p.
Subjects: Reinforcement learning, K-means clustering, Store location
Abstract: LBS services generate massive amounts of trajectory data over time, which will be shared with others for further intelligent services. Due to the ubiquity and openness of LBS, the user's trajectory may be collected and interfered by attackers. Existing solutions cannot take into account the privacy and availability of trajectory data at the same time. In order to achieve the balance between privacy and availability, the Hasse Diagram Sensitivity Differential Privacy with Reinforcement Learning (HDS-DPRL) is designed in this paper. The first module uses the ameliorated K-means clustering to reduce redundant position coordinate points. The second one includes an algorithm for calculating sensitive positions based on Hasse Diagram, which is used to store trajectories and construct a partial order relationship based on the access frequency of position point visits to calculate sensitive locations. The third module employs Reinforcement Learning to compute the optimal Laplace boundary and adds bounded Laplace noise to the Hasse Diagram that stores sensitive locations implement differential privacy. Extensive experiments on synthetic datasets and real-world datasets demonstrate the superior performance of our HDS-DPRL compared to the existing solutions while providing availability and privacy. Thus, HDS-DPRL can be applied to privacy-enhanced applications of trajectories. • The K-means algorithm is utilized to generalize trajectories to the same class. • Hasse Diagram is used to calculate sensitive locations. • Differential Privacy with Reinforcement Learning is designed for privacy-enhanced. • The performance of HDS-DPRL is analysed under different situations. [ABSTRACT FROM AUTHOR]
Copyright of Future Generation Computer Systems is the property of Elsevier B.V. and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database: Engineering Source
Description
Abstract:LBS services generate massive amounts of trajectory data over time, which will be shared with others for further intelligent services. Due to the ubiquity and openness of LBS, the user's trajectory may be collected and interfered by attackers. Existing solutions cannot take into account the privacy and availability of trajectory data at the same time. In order to achieve the balance between privacy and availability, the Hasse Diagram Sensitivity Differential Privacy with Reinforcement Learning (HDS-DPRL) is designed in this paper. The first module uses the ameliorated K-means clustering to reduce redundant position coordinate points. The second one includes an algorithm for calculating sensitive positions based on Hasse Diagram, which is used to store trajectories and construct a partial order relationship based on the access frequency of position point visits to calculate sensitive locations. The third module employs Reinforcement Learning to compute the optimal Laplace boundary and adds bounded Laplace noise to the Hasse Diagram that stores sensitive locations implement differential privacy. Extensive experiments on synthetic datasets and real-world datasets demonstrate the superior performance of our HDS-DPRL compared to the existing solutions while providing availability and privacy. Thus, HDS-DPRL can be applied to privacy-enhanced applications of trajectories. • The K-means algorithm is utilized to generalize trajectories to the same class. • Hasse Diagram is used to calculate sensitive locations. • Differential Privacy with Reinforcement Learning is designed for privacy-enhanced. • The performance of HDS-DPRL is analysed under different situations. [ABSTRACT FROM AUTHOR]
ISSN:0167739X
DOI:10.1016/j.future.2023.01.008