Social Network Privacy for Attribute Disclosure Attacks
- 1 July 2011
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 445-449
Abstract
Increasing research on social networks stresses the urgency for producing effective means of ensuring user privacy. Represented ubiquitously as graphs, social networks have a myriad of recently developed techniques to prevent identity disclosure, but the equally important attribute disclosure attacks have been neglected. To address this gap, we introduce an approach to anonymize social networks that have labeled nodes, α-proximity, which requires that the label distribution in every neighbourhood of the graph be close to that throughout the entire network. We present an effective greedy algorithm to achieve α-proximity and experimentally validate the quality of the solutions it derives.Keywords
This publication has 8 references indexed in Scilit:
- A New Approach to Manage Security against Neighborhood Attacks in Social Networks2010
- K-isomorphism2010
- k-symmetry model for identity anonymization in social networks2010
- Anonymizing bipartite graph data using safe groupingsThe VLDB Journal, 2009
- To join or not to join2009
- Towards identity anonymization on graphs2008
- Preserving Privacy in Social Networks Against Neighborhood Attacks2008
- L -diversityACM Transactions on Knowledge Discovery From Data, 2007