1. An efficient graph community clustering and ranking-based cryptographic privacy preserving framework on large social databases
- Author
-
Shamila. M, G. Rekha, and K. Vinuthna Reddy
- Subjects
Online social networking graph ,Privacy-Preserving Data Mining ,Multi-Privacy Control Framework ,community clustering ,cryptographic system ,Professor Swadesh Singh, Gokaraju Rangaraju Institute of Engineering and Technology, Hyderabad, India ,Engineering (General). Civil engineering (General) ,TA1-2040 - Abstract
In today’s digital age, individuals using online social platforms have created complex social connections, transforming the landscape of virtual relationships. In traditional graph-based models, various issues, such as community clustering, privacy preserving on large data and effective anomaly detection in each graph cluster, are unattended on large graph data. This work proposes a Multi-Online Social Networking Graph Data (MOSNGD) model to address crucial aspects of optimal community clustering, influence node detection and privacy-preserving techniques. Utilizing a graph-based model, this approach integrates an innovative optimal community clustering method (OCCM) to enhance the understanding of cohesive user groups with shared interests. Additionally, the introduced Top-k Influence Node Detection method (TINDM) aims to identify key nodes within these communities, facilitating the recognition of influential entities actively shaping the network’s dynamics. Recognizing the importance of privacy in the age of data-driven insights, proposed a Multi-Privacy Control Framework (MPCF) for Privacy-Preserving Data Mining (PPDM) on k-influence sensitive graph data. The experimental results illustrate momentous improvements when compared to existing methods across varied performance metrics, namely, (a) clustering accuracy, (b) influence node detection precision and (c) privacy preservation. Specifically, the proposed MOSNGD method attained (a) 98.9% clustering accuracy, (b) 99.1% precision rate in influence node detection and improved privacy protection rate when compared to traditional methods. The results highlight the efficacy of the proposed framework in addressing the key challenges in Online Social Network (OSN) analysis.
- Published
- 2024
- Full Text
- View/download PDF