Random walk with restart over dynamic graphs

Weiren Yu, Julie McCann

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

Abstract

Random Walk with Restart (RWR) is an appealing measure of proximity between nodes based on graph structures. Since real graphs are often large and subject to minor changes, it is prohibitively expensive to recompute proximities from scratch. Previous methods use LU decomposition and degree reordering heuristics, entailing O(|V|3) time and O(|V|2) memory to compute all (|V|2) pairs of node proximities in a static graph. In this paper, a dynamic scheme to assess RWR proximities is proposed: (1) For unit update, we characterize the changes to all-pairs proximities as the outer product of two vectors. We notice that the multiplication of an RWR matrix and its transition matrix, unlike traditional matrix multiplications, is commutative. This can greatly reduce the computation of all-pairs proximities from O(|V|3) to O(|Δ|) time for each update without loss of accuracy, where |Δ| (<<|V|2) is the number of affected proximities. (2) To avoid O(|V|2) memory for all pairs of outputs, we also devise efficient partitioning techniques for our dynamic model, which can compute all pairs of proximities segment-wisely within O(l|V|) is a user-controlled trade-off between  memory an I/O costs. (3) For bulk update, we also devise aggregation and hashing methods, which can discard many unneccessary updates further and handle chuncks of unit updates simultaneously. Our experimental results on various datasets demonstrate that our methods can be 1-2 orders of magnitude faster than other competitors while securing scalability and exactness...
Original languageEnglish
Title of host publicationProceedings: 16th IEEE International Conference on Data Mining, ICDM 2016
EditorsFrancesco Bonchi, Josep Domingo-Ferrer, Ricardo Baeza-Yates, et al
Place of PublicationPiscataway, NJ (US)
PublisherIEEE
Pages589-598
Number of pages10
ISBN (Electronic)978-1-5090-5473-2
ISBN (Print)978-1-5090-5474-9 , 978-1-5090-5472-5
DOIs
Publication statusPublished - 2017
Event16th IEEE International Conference on Data Mining and Workshops - World Trade Center, Bacelona, Spain
Duration: 12 Dec 201615 Dec 2016

Publication series

NameProceedings. IEEE International Conference on Data Mining
PublisherIEEE
ISSN (Print)2374-8486
ISSN (Electronic)2374-8486

Conference

Conference16th IEEE International Conference on Data Mining and Workshops
Abbreviated titleICDM &amp; ICDMW 2016
CountrySpain
CityBacelona
Period12/12/1615/12/16

Bibliographical note

© 2016 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.

Keywords

  • time measurement
  • loss measurement
  • Symmetric matrices
  • noise measurement
  • computational modeling
  • matrix decomposition
  • matrix converters

Fingerprint Dive into the research topics of 'Random walk with restart over dynamic graphs'. Together they form a unique fingerprint.

  • Cite this

    Yu, W., & McCann, J. (2017). Random walk with restart over dynamic graphs. In F. Bonchi, J. Domingo-Ferrer, R. Baeza-Yates, & et al (Eds.), Proceedings: 16th IEEE International Conference on Data Mining, ICDM 2016 (pp. 589-598). (Proceedings. IEEE International Conference on Data Mining). IEEE. https://doi.org/10.1109/ICDM.2016.0070