Fedivertex: a Graph Dataset based on Decentralized Social Media

Introduction
Social network graphs are central to graph learning research, serving as standard benchmarks for algorithm evaluation.

However, existing datasets focus mainly on mainstream social media platforms whose structures are shaped notably by algorithmic recommendations.

This raises an important question: would alternative, decentralized social networks exhibit different properties?
We address this by studying the Fediverse; a collection of decentralized social networks (such as Mastodon and Lemmy).

These platforms differ fundamentally from for-profit social media, notably in decentralization and absence of recommendation algorithms, which may yield distinct graph structures.
We introduce Fedivertex, a dataset of over 400 graphs from seven decentralized networks, collected weekly over more than a year.

The dataset, released with a companion Python package to facilitate its use, supports research on temporal and structural aspects of decentralized social networks.
Citation

This dataset is presented in a conference paper published at the ACM Web Conference 2026:  https://dl.acm.org/doi/10.1145/3774904.3792868
BibTeX entry: 

Python Package
We implemented a simple Python API to interact easily with the dataset: https://pypi.org/project/fedivertex/
pip3 install fedivertex
This package automatically downloads the dataset and generate NetworkX graphs:

from fedivertex import GraphLoader
loader.list_graph_types("mastodon")

# List available graphs for a given software, here federation and active_user
G = loader.get_graph(software = "mastodon", graph_type = "active_user", index = 0, only_largest_component = True)

# G contains the Networkx graph of the giant component of the active users graph at the 1st date of collection

Available graphs
The dataset contains graphs crawled on a daily basis on 7 social networks from the Fediverse. Each graph quantifies/characterizes the interaction differently depending on the information provided by the public API of these networks.
We present briefly the graph below (NB: the term "instance" refers to servers on the Fediverse):
- [Bookwyrm/Friendica/Lemmy/Misskey/Pleroma] "federation" graphs: If two instances know each other they are connected in this graph. The federation graph then corresponds to the undirected communication graph between instances.

- Peertube "follow" graphs: On Peertube, an instance X can follow an instance Y to let its users see all the videos posted on Y. This graph is a directed graph with edges of weight 1 for following.

- Lemmy "federation with blocks" graphs: This graph completes the federation graph with negative edges when an instance X blocks instance Y. The graph is directed.

- Lemmy "cross-instance" graphs: two instances are connected as soon as there exists a pair of users who publis
📤 Share this page
Found this useful? Share it with your network.
𝕏XfFacebookinLinkedIn✆WhatsApp✈Telegram✉EmailMMendeley⧉Copy LinkRGResearchGateAAcademia.edu
✓ Link copied! Paste it on ResearchGate / Academia.edu