Extraction of Community Transition Rules from Data Streams as Large Graph Sequence

In this study, we treat transactional sets of data streams as a graph sequence. This graph sequence represents both the relational structures of data for each period and changes in these structures. In addition, we analyze changes in a community in this graph sequence. Our proposed algorithm extract...

Full description

Saved in:
Bibliographic Details
Published inJournal of advanced computational intelligence and intelligent informatics Vol. 15; no. 8; pp. 1073 - 1081
Main Authors Yamaguchi, Takehiro, Niimi, Ayahiko
Format Journal Article
LanguageEnglish
Published 20.10.2011
Online AccessGet full text

Cover

More Information
Summary:In this study, we treat transactional sets of data streams as a graph sequence. This graph sequence represents both the relational structures of data for each period and changes in these structures. In addition, we analyze changes in a community in this graph sequence. Our proposed algorithm extracts community transition rules to detect communities that appear irregularly in a graph sequence using our proposed method combined with adaptive graph kernels and hierarchical clustering. In experiments using synthetic datasets and social bookmark datasets, we demonstrate that our proposed algorithm detects changes in a community appearing irregularly.
ISSN:1343-0130
1883-8014
DOI:10.20965/jaciii.2011.p1073