End user privacy preservation in social networks against neighborhood attack

Neighborhood attack over social networking sites breach the privacy of end users having an identical scenario. Recently numbers of anonymization techniques are developed to preserve privacy against neighborhood attack by adding dummy vertices and edges. Anonymization through dummy data increases the noise level in social networks data that lead to alter the originality and increase the information loss. This paper proposed one-neighborhood adjacency matrix (ONAM) base anonymization process in which optimum numbers of dummy edges are added into the unique sub-graph. The unique subgraph is suggested by 1-neighborhood adjacency matrix of their neighbors subgraph. In proposed anonymization process, the number of vertices remains same as that before anonymization. The addition of dummy edges increases isomorphic sub-networks and reduces the information loss. An analytic experiment is being conducted on Zachary's karate club, modified karate club, Les Miserables, and PolBooks four different datasets. After adding the dummy edges suggested by ONAM on these datasets, the noise level has measured as 6% to 8% with 0.0189 attack probability.