Focused SANA: Speeding up network alignment

Ilia Leybovich, Rami Puzis, Roni Stern, Maor Reuben

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Network Alignment (NA) is a generalization of the graph iso-morphism problem for non-isomorphic graphs, where the goal is to find a node mapping as close as possible to isomor-phism. Recent successful NA algorithms follow a search-based approach, such as simulated annealing. We propose to speed up search-based NA algorithms by pruning the search-space based on heuristic rules derived from the topological features of the aligned nodes. We define several desirable properties of such pruning rules, analyze them theoretically, and propose a pruning rule based on nodes' degrees. Experi-mental results show that using the proposed rule yields sig-nificant speedup and higher alignment quality compared to the state of the art. In addition, we redefine common NA ob-jective functions in terms of established statistical analysis metrics, opening a wide range of possible objective functions.

Original languageEnglish
Title of host publicationProceedings of the 11th International Symposium on Combinatorial Search, SoCS 2018
EditorsVadim Bulitko, Sabine Storandt
PublisherAAAI press
Pages168-172
Number of pages5
ISBN (Electronic)9781577358022
StatePublished - 1 Jan 2018
Event11th International Symposium on Combinatorial Search, SoCS 2018 - Stockholm, Sweden
Duration: 14 Jul 201815 Jul 2018

Publication series

NameProceedings of the 11th International Symposium on Combinatorial Search, SoCS 2018

Conference

Conference11th International Symposium on Combinatorial Search, SoCS 2018
Country/TerritorySweden
CityStockholm
Period14/07/1815/07/18

Fingerprint

Dive into the research topics of 'Focused SANA: Speeding up network alignment'. Together they form a unique fingerprint.

Cite this