logo Netzschleuder network catalog, repository, and centrifuge

telegram_channels_graph — Telegram channels graph (2015-2024)

Description

Networks of public Telegram channels, from the dataset of Shapovalova and Sofiienko, who collected 264,777 channels and their publications through the Telegram API. Each node is a channel, and there are two networks:

  • forwards: a directed edge from channel u to channel v means that publications of u were forwarded to v (v republished posts of u), with the number of forwarded publications as weight (2,207,614 edges, including 21,140 self-loops).
  • mentions: a directed edge from channel u to channel v means that publications of u mention v through a URL (a link to the channel or to one of its posts), with the number of mentions as weight (8,046,017 edges). The file contains repeated pairs of channels (933,000 of the edges duplicate a pair, with their own counts), which are kept as multiple edges, as they are.

The direction follows the columns FromId and ToId of the files (in forwards, e.g., the news agency RIA Novosti is the source of 133,600 forwarded publications, and the target of 45).

Both networks contain all channels of the dataset. For about half of them (133,603), the authors also give the 90th percentile of the views and of the forwards of their publications, and their main language (determined with fastText), at the time of a snapshot. A few channels (947) appear twice in the file of channels, with two main languages (ties), which are both kept. The mentions are made by 13,013 channels that are not in the file of channels; they are included as nodes without attributes (has_record false). The time range of the data is not stated; the channels were created from September 2015 to October 2024, and the snapshots were taken from 2015 to October 2024.

The networks have the following properties:

  • (Node) channel_id: the id of the channel on Telegram.
  • (Node) has_record: whether the channel is in the file of channels.
  • (Node) name: its username (its address is https://t.me/<name>; empty for some).
  • (Node) title.
  • (Node) created: the time of creation of the channel (Unix time).
  • (Node) snapshot: the time of the snapshot of the statistics (Unix time; -1 if there are none).
  • (Node) top_lang: the ISO code of the main language of its publications (two codes separated by a comma for ties; empty if unknown).
  • (Node) p90_views and p90_forwards: the 90th percentiles of the numbers of views and forwards of its publications at the time of the snapshot (NaN if unknown).
  • (Edge) count: the number of forwarded publications (forwards) or of mentions (mentions).
Tags
Social Online Communication Weighted Multigraph Metadata
Added
2026-10-31
Citation
Upstream URL OK
https://doi.org/10.5281/zenodo.14749767
Upstream license
Creative Commons Attribution 4.0 International (CC BY 4.0), https://creativecommons.org/licenses/by/4.0/
Networks
Tip: click on a column header to sort the list, or hover over it for a description.
Column legend
$\left<k\right>$
Mean degree
$\sigma_k$
Standard deviation of the degrees
$\lambda_h$
Largest eigenvalue of the Hashimoto (non-backtracking) matrix
$\tau$
Random walk mixing time
$r$
Degree assortativity
$c$
Global clustering coefficient
$\oslash$
(Pseudo-) diameter
$S$
Relative size of the largest component
NPs
Node properties
EPs
Edge properties
Name Nodes Edges $\left<k\right>$ $\sigma_k$ $\lambda_h$ $\tau$ $r$ $c$ $\oslash$ $S$ Kind Mode NPs EPs gt GraphML GML csv
forwards 264,777 2,207,614 8.34 72.28 417.50 741.98 0.04 0.10 22 0.90 Directed Unipartite channel_id has_record name title top_lang created snapshot p90_views p90_forwards count 17.6 MiB 27.4 MiB 26.6 MiB 26.0 MiB
mentions 277,790 8,046,017 28.96 570.62 2517.65 472.46 -0.45 0.02 23 0.44 Directed Unipartite channel_id has_record name title top_lang created snapshot p90_views p90_forwards count 11.3 MiB 31.6 MiB 30.3 MiB 39.8 MiB