Discograph

Discograph turns review data from ReviewDB into an interactive network that shows how people might be connected. It crawls from a seed Discord user ID, builds a graph, finds groups, and exports an HTML view you can open in a browser.

Features

• Fast async crawl from a seed user
• Clear NetworkX graph of connections
• Group detection with the Louvain algorithm
• Shareable HTML view
• Visuals that show clusters and bridges

Requirements

• Python 3.13+ (recommended)
• Poetry

Install dependencies via Poetry:

poetry install

Usage

Run from the project root:

poetry run python src --userid <DISCORD_USER_ID> --depth <CRAWL_DEPTH> --show-cross-communities

This will generate a file named discograph.html in the project directory. Open it in your browser.

Syntax help

• -cross or --show-cross-communities (Optional)

  • Shows interactions between communities.
  • Disable for large graphs to reduce lag and improve readability.

• --userid or -id

  • The root Discord user ID to start the crawl from.
  • Example: --userid 233958408179417089

• --depth or -d

  • How many layers to crawl outward from the root user.
  • Use small numbers (e.g., 1–3) for faster runs.
  • Example: --depth 2

Examples:

poetry run python src -id 233958408179417089 -d 2
poetry run python src --userid 704017415432044544 --depth 3 --show-cross-communities

Output

• discograph.html: interactive network view
  • Nodes colored by detected communities
  • Box nodes represent community hubs
  • Solid edges within communities, dashed gray edges between communities

Open discograph.html (if not done automatically) in any modern browser.

Feedback

If you are interested in social graphs or visualization, I would be glad to hear your thoughts and suggestions.