License: Creative Commons Attribution 4.0 International license (CC BY 4.0)
When quoting this document, please refer to the following
DOI: 10.4230/LIPIcs.GIScience.2023.15
URN: urn:nbn:de:0030-drops-189109
URL: http://dagstuhl.sunsite.rwth-aachen.de/volltexte/2023/18910/
Go to the corresponding LIPIcs Volume Portal

Ballatore, Andrea ; Cavazzi, Stefano

Why Is Greenwich so Common? Quantifying the Uniqueness of Multivariate Observations (Short Paper)

pdf-format:
LIPIcs-GIScience-2023-15.pdf (4 MB)

Abstract

The concept of uniqueness can play an important role when the assessment of an observation’s distinctiveness is essential. This article introduces a distance-based uniqueness measure that quantifies the relative rarity or commonness of a multi-variate observation within a dataset. Unique observations exhibit rare combinations of values, and not necessarily extreme values. Taking a cognitive psychological perspective, our measure defines uniqueness as the sum of distances between a target observation and all other observations. After presenting the measure u and its corresponding standardised version u_z, we propose a method to calculate a p value through a probability density function. We then demonstrate the measure’s behaviour in a case study on the uniqueness of Greater London boroughs, based on real-world socioeconomic variables. This initial investigation indicates that u can support exploratory data analysis.

BibTeX - Entry

@InProceedings{ballatore_et_al:LIPIcs.GIScience.2023.15,
  author =	{Ballatore, Andrea and Cavazzi, Stefano},
  title =	{{Why Is Greenwich so Common? Quantifying the Uniqueness of Multivariate Observations}},
  booktitle =	{12th International Conference on Geographic Information Science (GIScience 2023)},
  pages =	{15:1--15:6},
  series =	{Leibniz International Proceedings in Informatics (LIPIcs)},
  ISBN =	{978-3-95977-288-4},
  ISSN =	{1868-8969},
  year =	{2023},
  volume =	{277},
  editor =	{Beecham, Roger and Long, Jed A. and Smith, Dianna and Zhao, Qunshan and Wise, Sarah},
  publisher =	{Schloss Dagstuhl -- Leibniz-Zentrum f{\"u}r Informatik},
  address =	{Dagstuhl, Germany},
  URL =		{https://drops.dagstuhl.de/opus/volltexte/2023/18910},
  URN =		{urn:nbn:de:0030-drops-189109},
  doi =		{10.4230/LIPIcs.GIScience.2023.15},
  annote =	{Keywords: uniqueness, distinctiveness, similarity, outlier detection, multivariate data}
}

Keywords: uniqueness, distinctiveness, similarity, outlier detection, multivariate data
Collection: 12th International Conference on Geographic Information Science (GIScience 2023)
Issue Date: 2023
Date of publication: 07.09.2023
Supplementary Material: Software (R Code): https://github.com/andrea-ballatore/calculating-uniqueness

DROPS-Home | Fulltext Search | Imprint | Privacy Published by LZI