Spectral Modelling for Spatial Network Analysis

Pirouz Nourian, Samaneh Rezvani, Sevil Sariyildiz, Franklin van der Hoeven

Research output: Chapter in Book/Conference proceedings/Edited volumeConference contributionScientificpeer-review

192 Downloads (Pure)

Abstract

Spatial Networks represent the connectivity structure between units of space as a weighted graph whose links are weighted as to the strength of connections. In case of urban spatial networks, the units of space correspond closely to streets and in architectural spatial networks the units correspond to rooms, convex spaces or star-convex spaces. Once represented as a graph, a spatial network can be analysed using graph theory and spectral graph theory. We present four steps of modelling a spectrum for an urban spatial network; present an implementation of a state-ofthe-art spectral graph-drawing algorithm and showcase a Spatial Eigenvector Centrality index, which is based on a novel definition of spatial networks based on Fuzzy Closeness indicators computed using Easiest Path distances.
Original languageEnglish
Title of host publicationProceedings of the Symposium on Simulation for Architecture and Urban Design (simAUD 2016)
EditorsRamtin Attar, Angelos Chronis, Sean Hanna, Michela Turrin
PublishersimAUD
Pages103-110
Number of pages8
ISBN (Print)978-1-365-05872-1
Publication statusPublished - 2016
EventSimAUD EU 2016: 7th annual Symposium on Simulation for Architecture and Urban Design - London, United Kingdom
Duration: 16 May 201618 May 2016

Conference

ConferenceSimAUD EU 2016: 7th annual Symposium on Simulation for Architecture and Urban Design
Country/TerritoryUnited Kingdom
CityLondon
Period16/05/1618/05/16

Keywords

  • Spatial Network Analysis
  • Spectral Graph Theory
  • Spatial Eigenvector Centrality
  • Spectral Graph Drawing
  • Dominant Eigenvectors
  • Generalized Power Iteration

Fingerprint

Dive into the research topics of 'Spectral Modelling for Spatial Network Analysis'. Together they form a unique fingerprint.

Cite this