Improved OSN Privacy Preservation Approach using the Data Normalization Algorithm

The use of online social networks (OSNs) has increased dramatically in the wake of the COVID-19 pandemic, according to current data. OSNs are increasingly integral to the way of life for a lot of people. Consequently, security measures to ward off malevolent actors are becoming more necessary due to the growing reliance on ONSs. Any intruder trying to access ONS and steal sensitive user data, whether for profit or not, would be breaking the law. Consequently, it is critical to use OSNs to guarantee the various degrees of privacy. Assuring the k-anonymity privacy requirements in OSN remains a research problem, despite the introduction of several privacy preservation solutions at the network and user levels. Using clustering based on numerous network features, we provide a new approach to successfully anonymizing OSNs in this work. Edge, node, and user attribute privacy in OSN graphs is an objective that the clustering approach brings to light. Each cluster of the suggested model will be guaranteed k-anonymity according to this clustering method. In order to pre-process and improve the quality of raw OSN data, we first create the data normalization method. The next step in achieving k-anonymization is to partition the OSN data into distinct clusters based on various graph features. Using a real-world dataset and state-of-the-art methodologies, we assess the suggested approach. With respect to anonymization degree, information loss, and execution time, the suggested technique guarantees high-level privacy preservation when compared to state-of-the-art technologies.