This project analyzes the structural characteristics of social networks across three major online platforms:
The objective is to understand how users are connected on each platform, identify influential nodes, and detect communities within each network. Multiple community detection algorithms are applied and compared to highlight structural differences across platforms.
- Compute descriptive network statistics such as:
- Connectedness
- Degree and centrality measures
- Detect communities within each social network
- Compare community detection results across different algorithms
- Visualize network structures and detected communities
- Draw conclusions on structural similarities and differences between platforms
The analysis uses pre-processed network adjacency data stored in Excel files:
Facebook_Data.xlsxTwitter_Data.xlsxInstagram_Data.xlsx
Each file represents a network graph where nodes correspond to users and edges represent connections or interactions.
- Python
- numpy
- pandas
- warnings
- networkx
- networkx.algorithms.community
- Girvan–Newman algorithm
- Kernighan–Lin algorithm
- node2vec
- gensim
- scikit-learn
- adjusted_rand_score
- matplotlib