Accepted author manuscript, 1.22 MB, PDF document
Available under license: CC BY: Creative Commons Attribution 4.0 International License
Final published version
Licence: CC BY: Creative Commons Attribution 4.0 International License
Research output: Contribution to Journal/Magazine › Journal article › peer-review
Research output: Contribution to Journal/Magazine › Journal article › peer-review
}
TY - JOUR
T1 - Fuzzy Bayesian inference for mapping vague and place-based regions
T2 - a case study of sectarian territory
AU - Huck, Jonny
AU - Whyatt, Duncan
AU - Davies, Gemma
AU - Dixon, John
AU - Sturgeon, Brendan
AU - Hocking, Bree
AU - Tredoux, Colin
AU - Bryan, Dominic
PY - 2023/8/31
Y1 - 2023/8/31
N2 - The problem of mapping regions with socially-derived boundaries has been a topic of discussion in the GIS literature for many years. Fuzzy approaches have frequently been suggested as solutions, but none have been adopted. This is likely due to difficulties associated with determining suitable membership functions, which are often as arbitrary as the crisp boundaries that they seek to replace. This paper presents a novel approach to fuzzy geographical modelling that replaces the membership function with a possibility distribution that is estimated using Bayesian inference. In this method, data from multiple sources are combined to estimate the degree to which a given location is a member of a given set and the level of uncertainty associated with that estimate. The Fuzzy Bayesian Inference approach is demonstrated through a case study in which census data are combined with perceptual and behavioural evidence to model the territory of two segregated groups (Catholics and Protestants) in Belfast, Northern Ireland, UK. This novel method provides a robust empirical basis for the use of fuzzy models in GIS, and therefore has applications for mapping a range of socially-derived and otherwise vague boundaries.
AB - The problem of mapping regions with socially-derived boundaries has been a topic of discussion in the GIS literature for many years. Fuzzy approaches have frequently been suggested as solutions, but none have been adopted. This is likely due to difficulties associated with determining suitable membership functions, which are often as arbitrary as the crisp boundaries that they seek to replace. This paper presents a novel approach to fuzzy geographical modelling that replaces the membership function with a possibility distribution that is estimated using Bayesian inference. In this method, data from multiple sources are combined to estimate the degree to which a given location is a member of a given set and the level of uncertainty associated with that estimate. The Fuzzy Bayesian Inference approach is demonstrated through a case study in which census data are combined with perceptual and behavioural evidence to model the territory of two segregated groups (Catholics and Protestants) in Belfast, Northern Ireland, UK. This novel method provides a robust empirical basis for the use of fuzzy models in GIS, and therefore has applications for mapping a range of socially-derived and otherwise vague boundaries.
KW - Fuzzy
KW - Vague
KW - Place
KW - Segregation
KW - Territory
U2 - 10.1080/13658816.2023.2229894
DO - 10.1080/13658816.2023.2229894
M3 - Journal article
VL - 37
SP - 1765
EP - 1786
JO - International Journal of Geographical Information Science
JF - International Journal of Geographical Information Science
SN - 1365-8816
IS - 8
ER -