newsgroups_words — 20 Newsgroups word occurrences (1995)

Description

A bipartite network of Usenet newsgroup postings and the words they contain, from a reduced version of the "20 Newsgroups" collection prepared by S. Roweis. The 20 Newsgroups collection, assembled by K. Lang, consists of about 20,000 postings to 20 Usenet newsgroups. In this reduced version, the vocabulary is restricted to 100 selected words, and the data contain 16,242 postings, each containing at least one of them; each posting is connected to each of these words it contains (occurrence only, without counts). Each posting is labelled by the top-level hierarchy of its newsgroup (comp.*, rec.*, sci.* or talk.*). The data were obtained from the file "20news_w100.mat" formerly distributed on S. Roweis' data page, retrieved from the Internet Archive copy.

The network has the following properties:

(Node) is_word: whether the node is a word (otherwise a posting).
(Node) name: the word (empty for postings).
(Node) index (postings): the position of the posting in the source data (starting from 0).
(Node) newsgroup (postings): the top-level newsgroup hierarchy of the posting ("comp.*", "rec.*", "sci.*" or "talk.*").

Tags

Text Online Unweighted Metadata

Added

2026-10-25

Citation

S. Roweis, "Data for MATLAB hackers: 20 Newsgroups (20news_w100.mat)." web.archive.org
K. Lang, "NewsWeeder: Learning to Filter Netnews." Machine Learning Proceedings 1995, 331-339 (1995) 10.1016/B978-1-55860-377-6.50048-7 Sci-Hub

Upstream URL OK

https://cs.nyu.edu/~roweis/data.html

Networks

Tip: hover over a column header 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
newsgroups_words	16,342	65,451	8.01	64.48	77.29	4.58	-0.60	0.00	6	1.00	Undirected	Bipartite	is_word name index newsgroup		342 KiB	557 KiB	550 KiB	462 KiB

Ridiculograms^*

^* These are automatically generated force-directed visualizations, and can be quite meaningless for networks both big and small. They should not be taken seriously as sources of scientific insight. See here for a discussion.