Subgraph Matching Kernels for Attributed Graphs

Subgraph Matching Kernels for Attributed Graphs

Abstract

We propose graph kernels based on subgraphmatchings, i.e.structure-preserving bijec-tions between subgraphs. While recently pro-posed kernels based on common subgraphs(Wale et al., 2008; Shervashidze et al., 2009)in general can not be applied to attributedgraphs, our approach allows to rate mappingsof subgraphs by a flexible scoring schemecomparing vertex and edge attributes by ker-nels. We show that subgraph matching ker-nels generalize several known kernels. Tocompute the kernel we propose a graph-theoretical algorithm inspired by a classicalrelation between common subgraphs of twographs and cliques in their product graph ob-served by Levi (1973). Encouraging experi-mental results on a classification task of real-world graphs are presented.

Grafik Top
Authors
  • Kriege, Nils M.
  • Mutzel, Petra
Grafik Top
Shortfacts
Category
Paper in Conference Proceedings or in Workshop Proceedings (Paper)
Event Title
29th International Conference on Machine Learning (ICML)
Divisions
Data Mining and Machine Learning
Event Location
Edinburgh, Scotland, UK
Event Type
Conference
Event Dates
26.06.-01.07.2012
Series Name
ICML'12: Proceedings of the 29th International Coference on International Conference on Machine Learning
ISSN/ISBN
978-1-4503-1285-1
Publisher
icml.cc / Omnipress
Page Range
pp. 291-298
Date
26 June 2012
Official URL
http://icml.cc/2012/papers/542.pdf
Export
Grafik Top