This project focuses on analyzing social network data to understand various properties such as community structure, clustering coefficients, graph diameter, and privacy-preserving techniques.
- Data analysis with various metrics
- Anonymization techniques for privacy
- Community detection using Louvain method
- Visualization of graph properties
- Wiki-Vote
- Email-Eu-core
- Soc-Epinions1
- Facebook Combined
- p2p-Gnutella08
The analysis includes:
- Clustering Coefficient: Evaluating the density of connections within clusters of nodes.
- Graph Diameter: Assessing the longest shortest path between two nodes.
- NMI (Normalized Mutual Information): Comparing community structures before and after perturbations to evaluate privacy and utility.
- Clustering Entropy: Analyzing the variation in clustering coefficients to measure diversity and privacy.