citeseer_linqs — CiteSeer computer science papers (LINQS, 2003)

Description

A directed citation network of computer science papers, from the version of the CiteSeer dataset distributed by the LINQS group (Lu & Getoor; Sen et al.) as a benchmark for collective classification, also used (as one of the "Planetoid" datasets) in many works on graph neural networks. CiteSeer was a digital library and search engine for scientific papers, built by Giles, Bollacker and Lawrence by automatically crawling the web for papers and extracting their references. This dataset is a selection of 3,312 papers, each classified into one of six topics, chosen such that every paper cites or is cited by at least one other paper of the selection. Each paper is described by the presence or absence of 3,703 words of a vocabulary obtained after stemming and removing stopwords, and discarding words that appear in fewer than 10 papers. The words themselves are not included in the data. This is different data from the one in the citeseer entry; see also the cora_linqs and pubmed_linqs entries.

An edge from paper A to paper B means that A cites B. The citation list refers to 15 papers that are not in the list of papers (17 citations); they are included as nodes without topic and words. The data contain 124 self-loops (papers citing themselves), which are kept.

The network has the following properties:

id: the identifier of the paper in the source data.
subject: the topic of the paper, one of Agents, AI (artificial intelligence), DB (databases), HCI (human-computer interaction), IR (information retrieval) and ML (machine learning), or empty for the papers missing from the list of papers.
words: a vector of 3,703 binary values, where the i-th value is 1 if the i-th word of the vocabulary appears in the paper, and 0 otherwise (empty for the papers missing from the list of papers).

Tags

Informational Citation Unweighted Metadata Text

Citation

C. L. Giles, K. D. Bollacker, S. Lawrence, "CiteSeer: An Automatic Citation Indexing System." Proceedings of the Third ACM Conference on Digital Libraries, 89-98 (1998) 10.1145/276675.276685 Sci-Hub
Q. Lu, L. Getoor, "Link-based Classification." Proceedings of the Twentieth International Conference on Machine Learning (ICML), 496-503 (2003) cdn.aaai.org
P. Sen, G. Namata, M. Bilgic, L. Getoor, B. Gallagher, T. Eliassi-Rad, "Collective Classification in Network Data." AI Magazine 29(3), 93-106 (2008) 10.1609/aimag.v29i3.2157 Sci-Hub

Upstream URL [Errno 111] Connection refused

https://linqs.org/datasets/#citeseer-doc-classification
https://web.archive.org/web/20260923154413/https://linqs-data.soe.ucsc.edu/public/lbc/

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
citeseer_linqs	3,327	4,732	1.42	3.42	11.57	662.53	0.05	0.13	28	0.64	Directed	Unipartite	id subject words		221 KiB	333 KiB	333 KiB	53 KiB