Data Anonymization in Social Networks State of the Art, Exposure of Shortcomings and Discussion of New Innovations

Privacy is a concern of social network users. Social networks are a source of valuable data for scientific or commercial analysis. Therefore, anonymizing social network data before releasing it becomes an important issue. The nodes in the network represent the individuals and the links among them denote their relationships. Nevertheless, publishing a social graph directly by simply removing the names of people who contributed to this graph raises important privacy issues. In particular, some inference attacks on the published graph can lead to de-anonymizing certain nodes, learning the existence of a social relation between two nodes or even using the structure of the graph itself to deduce the value of certain sensitive attributes. In this paper, we present a brief yet systematic review of the existing anonymization techniques for privacy preserving publishing of social network data. We identify the challenges in privacy preserving publishing of social network data comparing to the extensively studied relational case. We survey the existing anonymization methods for privacy preservation in three categories: graph modification approaches, generalization approaches and differential privacy methods.