logo Netzschleuder network catalog, repository, and centrifuge

wikilens_ratings — WikiLens ratings and links (2004-2008)

Description

A bipartite network of the users of WikiLens and the items they rated, and the network of links among its wiki pages, from the database dump that GroupLens (University of Minnesota) published on 2 February 2008. WikiLens was a research project of GroupLens, a generalized, community-maintained recommender system: a wiki (based on PhpWiki) on which users could define item types (e.g. beers, books, restaurants or films) and categories, add items, and rate them and get recommendations. It ran from 2004 and was taken offline in 2009. The dump contains all the data that could be obtained by crawling the site; private data (user accounts, preferences, logs) were erased from it.

The network ratings is bipartite: an edge connects a user to an item they rated, for all 26,937 ratings of the dump, with the rating (from 0 to 5, in half stars) and its time. Each user is identified by their user page and each rated object by its URN; most URNs are those of wiki pages (an item and its page), but a few name other objects (e.g. names in a list or titles of papers), and 10 ratings refer to URN ids that are absent from the dump, whose items have no name. Users are also wiki pages, which can themselves be rated as items; such a page is then both a user node and an item node. The times are given in the dump without a time zone, and are taken to be the local time of the University of Minnesota (US Central time, America/Chicago, with daylight saving time).

The network links is directed: nodes are wiki pages and an edge $i \to j$ means that page $i$ links to page $j$, from the link cache of the wiki (7,430 links, including self-links). Only the pages with links are included; 9 links refer to page ids absent from the dump, whose nodes have no name. Many of the dump's other pages are spam or empty, and are left out.

The network ratings has the following properties:

  • (Node) name: the name of the user page, or of the item (its wiki page name, or its URN if it is not a wiki page; empty if unknown).
  • (Node) is_user: whether the node is a user (otherwise an item).
  • (Node) urn: the URN of the item (empty for users).
  • (Node) category: the categories (item types) of the page of the item or user (e.g. Movie, Book, Beer, User).
  • (Edge) rating: the rating, from 0 to 5.
  • (Edge) dimension: the dimension of the rating (0 for all ratings but one, of dimension 5).
  • (Edge) item_version: the version of the item's wiki page that was rated (-1 if unknown).
  • (Edge) time: the time of the rating, in Unix time.

The network links has the following properties:

  • (Node) name: the name of the wiki page (empty if unknown).
  • (Node) category: the categories of the page.
  • (Node) hits: the number of visits of the page recorded by the wiki.
Tags
Social Online Preferences Bipartite Weighted Timestamps Metadata
Added
2026-10-27
Citation
  • D. Frankowski, S. K. Lam, S. Sen, F. M. Harper, S. Yilek, M. Cassano, J. Riedl, "Recommenders everywhere: the WikiLens community-maintained recommender system." Proceedings of the 2007 International Symposium on Wikis (WikiSym), 47-60 (2007). 10.1145/1296951.1296957 Sci-Hub
Upstream URL OK
https://grouplens.org/datasets/wikilens/
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
ratings 5,437 26,937 9.91 47.29 68.96 7.30 -0.34 0.00 9 1.00 Undirected Bipartite name is_user urn category rating dimension item_version time 288 KiB 437 KiB 428 KiB 377 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.