Title: Research on personalised privacy-preserving model of multi-sensitive attributes

Authors: Haiyan Kang; Yaping Feng; Xiameng Si; Kaili Lu

Addresses: School of Information and Management, Beijing Information Science and Technology University, Beijing 100192, China ' School of Information and Management, Beijing Information Science and Technology University, Beijing 100192, China ' School of Information and Management, Beijing Information Science and Technology University, Beijing 100192, China ' School of Computer, Beijing Information Science and Technology University, Beijing 100192, China

Abstract: In order to protect user information from being leaked, it is imperative to improve the availability of published data and realise the safe and efficient information sharing. Aiming at the anonymous privacy-preserving of multi-sensitive attribute data release in logistics industry, this paper proposes a personalised privacy-preserving model of multi-sensitive attributes with weights clustering and dividing (PMSWCD) by analysing existing model. Firstly, according to the different needs of users, the corresponding weight is set for each sensitive attribute value to realise personalisation and then weighted clustering. Secondly, divide the records according to the weighted average value, and select records to establish a group that satisfies l-diversity. Finally, release data based on the idea of multi-dimensional bucket. Through experimental analysis, compared with WMBF algorithm, the release ratio of important data of PMSWCD algorithm proposed in this paper is significantly improved, reaching more than 95%, which improves the availability of data.

Keywords: multi-sensitive attributes; data release; personalised; privacy-preserving; weights clustering; dividing; multi-dimensional bucket; l-diversity.

DOI: 10.1504/IJIPT.2023.129749

International Journal of Internet Protocol Technology, 2023 Vol.16 No.1, pp.58 - 67

Received: 09 Feb 2021
Accepted: 21 Aug 2021
Published online: 23 Mar 2023 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article