Protecting private geosocial networks against practical hybrid attacks with heterogeneous information

Publisher:
Elsevier
Publication Type:
Journal Article
Citation:
Neurocomputing, 2016, 210 pp. 81 - 90
Issue Date:
2016-06-14
Full metadata record
Files in This Item:
Filename Description Size
1-s2.0-S0925231216305896-main.pdfPublished Version2.11 MB
Adobe PDF
© 2016 Elsevier B.V.GeoSocial Networks (GSNs) are becoming increasingly popular due to its power in providing high-performance and flexible service capabilities. More and more Internet users have accepted this innovative service model. However, even GSNs have great business value for data analysis by integrated with location information, it may seriously compromise users' privacy in publishing the GSN data. In this paper, we study the identity disclosure problem in publishing GSN data. We first discuss the attack problem by considering both the location-based and structure-based properties, as background knowledge, and then formalize two general models, named (k,m)-anonymity and (k,m,l)-anonymity Then we propose a complete solution to achieve (k,m)-anonymization and (k,m,l)-anonymization to prevent the released data from the above attacks above. We also take data utility into consideration by defining specific information loss metrics. It is validated by real-world data that the proposed methods can prevent GSN dataset from the attacks while retaining good utility.
Please use this identifier to cite or link to this item: