A novel clustering-based anonymization approach for graph to achieve Privacy Preservation in Social Network

Abstract

Abstract. Serious privacy concern rises with the prosperity of social network applications. To prevent the privacy of vertices or edges associated with entities in a social network from getting re-identified through background information or queries,a novel clustering-based approach is proposed to anonymize vertices and edges. Concepts of vertex similarity matrix and the distance between a vertex and a cluster are defined, based on which a k -anonymized graph approach is presented. The effectiveness of the approach is verified Through experiments that compare the performance of our method with that of SASN, an existing anonymization algorithm .

Privacy preservation for social networks sequential publishing

“…In [14], the case of labeled nodes has been considered, so the clusters are created, based on attributes and neighborhood similarity. Similar anonymization schemes are used in [18]. As said, authors in [18] propose a solution that first divides the graph of n nodes into several clusters; and each (cluster) contains at least k nodes.…”

“…Similar anonymization schemes are used in [18]. As said, authors in [18] propose a solution that first divides the graph of n nodes into several clusters; and each (cluster) contains at least k nodes. Then they replace the subgraph, which contains similar nodes into a cluster, by a super node.…”

“…However, the full publication of these social network data violates the users' privacy because an adversary can infer the affiliation links that the victim would like to keep private. To overcome this problem, researchers have proposed several techniques [3] [9] [18][20] to anonymize the data before their publication so that the privacy of users is preserved and the needs of data miners are satisfied. These techniques deal with different privacy risks which are [26]:…”

A Novel Graph-modification Technique for User Privacy-preserving on Social Networks

“…Many researchers have focused on studying the problem of user privacy in online social networks [4]- [8]. To improve the security of social networks, various social network data anonymization techniques have been proposed [9]- [15]. These anonymization approaches are designed based on the principle of k-anonymity.…”

Privacy Preservation in Social Networks Sequential Publishing

“…However, the full publication of these social network data violates the users' privacy because an adversary can infer the affiliation links that the victim would like to keep private. To overcome this problem, researchers have proposed several techniques [3] [9] [18] [20] to anonymize the data before their publication so that the privacy of users is preserved and the needs of data miners are satisfied. These techniques deal with different privacy risks which are [26]:…”