Learning to rank graphs for online similar graph search

Bingjun Sun, Prasenjit Mitra, C. Lee Giles

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

2 Citations (Scopus)

Abstract

Many applications in structure matching require the ability to search for graphs that are similar to a query graph, i.e., similarity graph queries. Prior works, especially in chemoinformatics, have used the maximum common edge subgraph (MCEG) to compute the graph similarity. This approach is prohibitively slow for real-time queries. In this work, we propose an algorithm that extracts and indexes subgraph features from a graph dataset. It computes the similarity of graphs using a linear graph kernel based on feature weights learned offline from a training set generated using MCEG. We show empirically that our proposed algorithm of learning to rank graphs can achieve higher normalized discounted cumulative gain compared with existing optimal methods based on MCEG. The running time of our algorithm is orders of magnitude faster than these existing methods.

Original languageEnglish
Title of host publicationInternational Conference on Information and Knowledge Management, Proceedings
Pages1871-1874
Number of pages4
DOIs
Publication statusPublished - 2009
Externally publishedYes
EventACM 18th International Conference on Information and Knowledge Management, CIKM 2009 - Hong Kong
Duration: 2 Nov 20096 Nov 2009

Other

OtherACM 18th International Conference on Information and Knowledge Management, CIKM 2009
CityHong Kong
Period2/11/096/11/09

Fingerprint

Graph
Learning to rank
Query
Kernel

Keywords

  • Graph kernel
  • Learn to rank
  • Similarity graph search

ASJC Scopus subject areas

  • Business, Management and Accounting(all)
  • Decision Sciences(all)

Cite this

Sun, B., Mitra, P., & Giles, C. L. (2009). Learning to rank graphs for online similar graph search. In International Conference on Information and Knowledge Management, Proceedings (pp. 1871-1874) https://doi.org/10.1145/1645953.1646252

Learning to rank graphs for online similar graph search. / Sun, Bingjun; Mitra, Prasenjit; Giles, C. Lee.

International Conference on Information and Knowledge Management, Proceedings. 2009. p. 1871-1874.

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

Sun, B, Mitra, P & Giles, CL 2009, Learning to rank graphs for online similar graph search. in International Conference on Information and Knowledge Management, Proceedings. pp. 1871-1874, ACM 18th International Conference on Information and Knowledge Management, CIKM 2009, Hong Kong, 2/11/09. https://doi.org/10.1145/1645953.1646252
Sun B, Mitra P, Giles CL. Learning to rank graphs for online similar graph search. In International Conference on Information and Knowledge Management, Proceedings. 2009. p. 1871-1874 https://doi.org/10.1145/1645953.1646252
Sun, Bingjun ; Mitra, Prasenjit ; Giles, C. Lee. / Learning to rank graphs for online similar graph search. International Conference on Information and Knowledge Management, Proceedings. 2009. pp. 1871-1874
@inproceedings{a6414597f02e4b16af36ee32d6355243,
title = "Learning to rank graphs for online similar graph search",
abstract = "Many applications in structure matching require the ability to search for graphs that are similar to a query graph, i.e., similarity graph queries. Prior works, especially in chemoinformatics, have used the maximum common edge subgraph (MCEG) to compute the graph similarity. This approach is prohibitively slow for real-time queries. In this work, we propose an algorithm that extracts and indexes subgraph features from a graph dataset. It computes the similarity of graphs using a linear graph kernel based on feature weights learned offline from a training set generated using MCEG. We show empirically that our proposed algorithm of learning to rank graphs can achieve higher normalized discounted cumulative gain compared with existing optimal methods based on MCEG. The running time of our algorithm is orders of magnitude faster than these existing methods.",
keywords = "Graph kernel, Learn to rank, Similarity graph search",
author = "Bingjun Sun and Prasenjit Mitra and Giles, {C. Lee}",
year = "2009",
doi = "10.1145/1645953.1646252",
language = "English",
isbn = "9781605585123",
pages = "1871--1874",
booktitle = "International Conference on Information and Knowledge Management, Proceedings",

}

TY - GEN

T1 - Learning to rank graphs for online similar graph search

AU - Sun, Bingjun

AU - Mitra, Prasenjit

AU - Giles, C. Lee

PY - 2009

Y1 - 2009

N2 - Many applications in structure matching require the ability to search for graphs that are similar to a query graph, i.e., similarity graph queries. Prior works, especially in chemoinformatics, have used the maximum common edge subgraph (MCEG) to compute the graph similarity. This approach is prohibitively slow for real-time queries. In this work, we propose an algorithm that extracts and indexes subgraph features from a graph dataset. It computes the similarity of graphs using a linear graph kernel based on feature weights learned offline from a training set generated using MCEG. We show empirically that our proposed algorithm of learning to rank graphs can achieve higher normalized discounted cumulative gain compared with existing optimal methods based on MCEG. The running time of our algorithm is orders of magnitude faster than these existing methods.

AB - Many applications in structure matching require the ability to search for graphs that are similar to a query graph, i.e., similarity graph queries. Prior works, especially in chemoinformatics, have used the maximum common edge subgraph (MCEG) to compute the graph similarity. This approach is prohibitively slow for real-time queries. In this work, we propose an algorithm that extracts and indexes subgraph features from a graph dataset. It computes the similarity of graphs using a linear graph kernel based on feature weights learned offline from a training set generated using MCEG. We show empirically that our proposed algorithm of learning to rank graphs can achieve higher normalized discounted cumulative gain compared with existing optimal methods based on MCEG. The running time of our algorithm is orders of magnitude faster than these existing methods.

KW - Graph kernel

KW - Learn to rank

KW - Similarity graph search

UR - http://www.scopus.com/inward/record.url?scp=74549114188&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=74549114188&partnerID=8YFLogxK

U2 - 10.1145/1645953.1646252

DO - 10.1145/1645953.1646252

M3 - Conference contribution

SN - 9781605585123

SP - 1871

EP - 1874

BT - International Conference on Information and Knowledge Management, Proceedings

ER -