Network Structure, Temporal Evolution and Non-protagonist Hubs in The Simpsons

Course: 02805 Social Graphs and Interactions
Institution: Danmarks Tekniske Universitet (DTU)
Authors: Yashasvi Tirpur Vishwanath, Chinmay Prasad Dongarkar, Vaibhav Anil Nagwani

Project Overview

This project analyzes the network structure of 104 secondary characters from The Simpsons using wiki-based co-appearance data from 753 episodes. We examine:

1. Community Detection: How secondary characters cluster into social and thematic groups
2. Temporal Evolution: How network structure changes across seasons
3. Non-protagonist Hubs: Which characters emerge as structural connectors when central nodes are removed
4. Text-Network Alignment: Relationship between structural communities and semantic similarity

Key Findings

• Network Properties: 104 nodes, 161 edges, density 0.030, 32 connected components
• Community Structure: 38 communities detected with modularity Q = 0.420
• Small World Properties: Diameter of 6, average path length 2.84, clustering coefficient 0.27
• Text-Network Alignment: Characters in the same community show significantly higher semantic similarity (p < 0.001)

Repository Structure

├── Group25_simpsons_Project_B.ipynb          # Main analysis notebook
├── Figures/                      # Generated figures
│   ├── figure1_network_overview.png
│   ├── figure2_centrality_comparison.png
│   ├── figure3_temporal_evolution.png
│   └── figure4_non_protagonist_hubs.png
├── character_centralities.csv    # Centrality measures
├── character_centrality_comparison.csv
└── character_temporal_trajectories.csv

Requirements

pip install networkx pandas numpy matplotlib seaborn scikit-learn nltk scipy

Methods

• Network Construction: Bipartite character-episode network projected to character-character co-appearance network
• Community Detection: Louvain algorithm
• Centrality Measures: Degree, betweenness, PageRank, eigenvector, closeness
• Text Analysis: TF-IDF vectorization and cosine similarity
• Temporal Analysis: Season-based network slices across 5 eras

Results

The analysis reveals:

• Strong modular organization with communities aligned to narrative settings (schools, workplaces, etc.)
• Temporal evolution toward specialization in later seasons
• Structural robustness with multiple alternative hubs
• Significant alignment between network structure and textual themes

References

• Newman, M. E. J. (2018). Networks: An Introduction (2nd ed.). Oxford University Press.
• Barabási, A.-L. (2016). Network Science. Cambridge University Press.
• Blondel, V. D., et al. (2008). Fast unfolding of communities in large networks. Journal of Statistical Mechanics.
• Data source: Simpsons Wiki

Authors

• Yashasvi Tirpur Vishwanath (s242563)
• Chinmay Prasad Dongarkar (s250155)
• Vaibhav Anil Nagwani (s242572)

License

This project is for academic purposes as part of the 02805 Social Graphs and Interactions course at DTU.

Acknowledgments

Data collected from Wikisimpsons (simpsons.fandom.com). Code and visualizations were refined with technical assistance.