Categorising neighbourhoods using OpenStreetMap POIs: Affinity propagation clustering of 7,213 subdistricts in Thailand
Issued Date
2024-01-01
Resource Type
ISSN
22265856
eISSN
25890360
Scopus ID
2-s2.0-85210083280
Journal Title
Journal of Urban Management
Rights Holder(s)
SCOPUS
Bibliographic Citation
Journal of Urban Management (2024)
Suggested Citation
Taecharungroj V., Ntounis N. Categorising neighbourhoods using OpenStreetMap POIs: Affinity propagation clustering of 7,213 subdistricts in Thailand. Journal of Urban Management (2024). doi:10.1016/j.jum.2024.11.005 Retrieved from: https://repository.li.mahidol.ac.th/handle/20.500.14594/102232
Title
Categorising neighbourhoods using OpenStreetMap POIs: Affinity propagation clustering of 7,213 subdistricts in Thailand
Author(s)
Author's Affiliation
Corresponding Author(s)
Other Contributor(s)
Abstract
The effective categorisation of neighbourhoods is a critical component of urban planning and development, providing a systematic framework for identifying and addressing the distinct characteristics and needs of different areas. This study utilises data from the open-source platform OpenStreetMap (OSM) to propose a novel approach to neighbourhood categorisation, with a focus on amenities as key elements. Data were collected on 4,121,900 points of interest (POIs) across 7213 subdistricts in Thailand, and the categorisation was conducted using the Affinity Propagation (AP) clustering technique. Through this approach, ten distinct neighbourhood clusters in Thailand were identified, demonstrating the efficacy of integrating OSM data with AP clustering. The findings underscore the necessity for more evidence-based planning policies aimed at enhancing amenities, vibrancy, and overall quality of life in neighbourhoods by promoting innovation and the development of creative districts. Furthermore, the study advocates for the consideration of ecological urbanism as an alternative pathway for neighbourhood development, a concept that has yet to be thoroughly explored in Thailand.