/fediverse-explorer

Visualization and Exploration of the Fediverse

Primary LanguageJupyter NotebookMIT LicenseMIT

Enter the Fediverse Explorer

This is a small personal project to visualize the Fediverse.

The Viewer shows a point cloud of Mastodon instances. The embedding is done with a spring model, where peering instances are placed close to each other.

How to use

  1. python pipeline/instance_scraper.py to create the SQLite database with all instances and peers.
    • This will take a while, and can be interrupted and resumed.
  2. python pipeline/add_index.py to add an index to the database. This will speed up the next steps
  3. python pipeline/embed_instances.py to embed the instances in a 2D space. This will take a while (few hours on my machine).
    • creates a graph.bin file, which is a binary file containing the graph of the instances.
    • creates a graph_mapping.json file, which is a mapping from instance id to index in the graph.
    • creates an embedding.json file.