Efficient methods and readily customizable libraries for managing complexity of large networks

Simple item page
buir.contributor.authorDoğrusöz, Uğur
buir.contributor.authorKaraçelik, Alper
buir.contributor.authorSafarli, İlkin
buir.contributor.authorBalcı, Hasan
buir.contributor.authorDervishi, Leonard
buir.contributor.authorSiper, Metin Can
dc.citation.epage18en_US
dc.citation.issueNumber5en_US
dc.citation.spage1en_US
dc.citation.volumeNumber13en_US
dc.contributor.authorDoğrusöz, Uğuren_US
dc.contributor.authorKaraçelik, Alperen_US
dc.contributor.authorSafarli, İlkinen_US
dc.contributor.authorBalcı, Hasanen_US
dc.contributor.authorDervishi, Leonarden_US
dc.contributor.authorSiper, Metin Canen_US
dc.date.accessioned2019-02-21T16:08:06Z
dc.date.available2019-02-21T16:08:06Z
dc.date.issued2018en_US
dc.departmentDepartment of Computer Engineeringen_US
dc.description.abstractBackground One common problem in visualizing real-life networks, including biological pathways, is the large size of these networks. Often times, users find themselves facing slow, non-scaling operations due to network size, if not a “hairball” network, hindering effective analysis. One extremely useful method for reducing complexity of large networks is the use of hierarchical clustering and nesting, and applying expand-collapse operations on demand during analysis. Another such method is hiding currently unnecessary details, to later gradually reveal on demand. Major challenges when applying complexity reduction operations on large networks include efficiency and maintaining the user’s mental map of the drawing. Results We developed specialized incremental layout methods for preserving a user’s mental map while managing complexity of large networks through expand-collapse and hide-show operations. We also developed open-source JavaScript libraries as plug-ins to the web based graph visualization library named Cytsocape.js to implement these methods as complexity management operations. Through efficient specialized algorithms provided by these extensions, one can collapse or hide desired parts of a network, yielding potentially much smaller networks, making them more suitable for interactive visual analysis. Conclusion This work fills an important gap by making efficient implementations of some already known complexity management techniques freely available to tool developers through a couple of open source, customizable software libraries, and by introducing some heuristics which can be applied upon such complexity management techniques to ensure preserving mental map of users.
dc.description.provenanceMade available in DSpace on 2019-02-21T16:08:06Z (GMT). No. of bitstreams: 1 Bilkent-research-paper.pdf: 222869 bytes, checksum: 842af2b9bd649e7f548593affdbafbb3 (MD5) Previous issue date: 2018en
dc.identifier.doi10.1371/journal.pone.0197238
dc.identifier.issn1932-6203
dc.identifier.urihttp://hdl.handle.net/11693/50398
dc.language.isoEnglish
dc.publisherPublic Library of Science
dc.relation.isversionofhttps://doi.org/10.1371/journal.pone.0197238
dc.rightsinfo:eu-repo/semantics/openAccess
dc.source.titlePLoS ONEen_US
dc.titleEfficient methods and readily customizable libraries for managing complexity of large networksen_US
dc.typeArticleen_US

Files

Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
Efficient_method_and_readily_customizable.pdf
Size:
7.27 MB
Format:
Adobe Portable Document Format
Description:
Full printable version
Download

Collections

Department of Computer Engineering