Wikipedia Clustering

Features

• Can be used with graph of links as descriptors : AnalyzerLinks program
• Can be used with content of articles as descriptors : AnalyzerContent program
• Can parse wikipedia dumps
• Html display of results
• Uses all libraries (clustering library, string B-Tree, trie library, tokenizer library)
• Can also be used to find all articles related with a specific article (using a similarity measure and a similarity threshold)
• You can use it for non-commercial and commercial purposes, licensed under terms of GNU LGPL license

Download

Download WikiClust-0.1.tgz

Some Results for Clustering of links between articles

(Without forcing each document to enter into a cluster)

	Similarity Threshold: 0.5	Similarity threshold: 0.33	Similarity threshold: 0.25
English	Here	Here	Here
Deutsch	Here	Here	Here
Français	Here	Here	Here
Polski	Here	Here	Here
日本語	Here	Here	Here
Nederlands	Here	Here	Here
Italiano	Here	Here	Here
Svenska	Here	Here	Here
Português	Here	Here	Here
Español	Here	Here	Here

Results for Clustering of articles content

(Without forcing each document to enter into a cluster)

	Similarity Threshold: 0.5	Similarity threshold: 0.33	Similarity threshold: 0.25
Français	Here	Here	Here
Polski	Here	Here	Here
日本語	Here	Here	Here
Nederlands	Here	Here	Here
Italiano	Here	Here	Here
Svenska	Here	Here	Here
Português	Here	Here	Here
Español	Here	Here	Here