A survey on privacy-preserving techniques for social network data

Abstract

In this era of 20th century, online social networks such as Facebook and Twitter play a very important role in everyone’s life. Social network data, regarding any individual organization, can be published online at any time, in which there is a risk of information leakage of anyone’s personal data. Hence, preserving the privacy of individual organizations and companies is needed before data are published online. Therefore, this research was carried out in this area for many years and it is still going on. There have been various existing techniques that provide the solutions for preserving privacy to tabular data called as relational data and also social network data represented in graphs. Different techniques exist for tabular data but we cannot apply directly to the structured complex graph data, which consist of vertices represented as individuals and edges represented as some kind of connection or relationship between the nodes. Various techniques such as K-anonymity, L-diversity, and T-closeness exist to provide privacy to nodes, and techniques such as edge perturbation and edge randomization are available to provide privacy to edges in social network graphs. Development of new techniques by integration into the exiting techniques such as K-anonymity, edge perturbation, edge randomization, and L-diversity to provide more privacy to relational data and social network data is ongoing in the best possible manner.