Extraction of fuzzy clusters from weighted graphs

Seiji Hotta, Kohei Inoue, Kiichi Urahama

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)


A spectral graph method is presented for partitioning of nodes in a graph into fuzzy clusters on the basis of weighted adjacency matrices. Extraction of a fuzzy cluster from a node set is formulated by an eigenvalue problem and clusters are extracted sequentially from major one to minor ones. A clustering scheme is devised at first for undirected graphs and it is next extended to directed graphs and also to undirected bipartite ones. These clustering methods are applied to analysis of a link structure in Web networks and image retrieval queried by keywords or sample images. Extracted structure of clusters is visualized by a multivariate exploration method called the correspondence analysis.

Original languageEnglish
Title of host publicationKnowledge Discovery and Data Mining
Subtitle of host publicationCurrent Issues and New Applications - 4th Pacific-Asia Conference, PAKDD 2000, Proceedings
EditorsTakao Terano, Huan Liu, Arbee L.P. Chen
PublisherSpringer Verlag
Number of pages12
ISBN (Print)3540673822, 9783540673828
Publication statusPublished - 2000
Event4th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2000 - Kyoto, Japan
Duration: Apr 18 2000Apr 20 2000

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Other4th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2000

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • General Computer Science


Dive into the research topics of 'Extraction of fuzzy clusters from weighted graphs'. Together they form a unique fingerprint.

Cite this