Airbnb, hotels, and saturation of the food industry: A multi-scale GWR approach

This paper evaluates the relationship between the food industry and the local variations of temporary accommodation (TAs, including hotels and short-term rentals). The aim is to capture the variance of the local statistic and pinpoint areas where food and beverages (F&B) presence is highly relat...

Full description

Saved in:
Bibliographic Details
Published in:arXiv.org 2019-06
Main Authors: Shabrina, Zahratu, Buyuklieva, Boyana, Matthew Ng Kok Ming
Format: Article
Language:English
Subjects:
Beverages
Food processing industry
Hotels
Hotels & motels
Multiscale analysis
Regression analysis
Regression models
Residential areas
Spatial analysis
Statistical analysis
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This paper evaluates the relationship between the food industry and the local variations of temporary accommodation (TAs, including hotels and short-term rentals). The aim is to capture the variance of the local statistic and pinpoint areas where food and beverages (F&B) presence is highly related to TAs in London. We explain the phenomena using OLS and compare the result with the local model - Geographically Weighted Regression (GWR) and multi-scale GWR (Fotheringham et al., 2017) allowing the use of different optimal bandwidths instead of assuming that relationship varies at the same spatial scale. The comparison is presented and the result shows that the GWR model shows significant improvement over Ordinary Least Square (OLS), increasing the R-squared from 0.28 to 0.75. MGWR further improves the model estimate, increasing the R-squared to 0.77, indicating the relationship happens in different spatial scales. Lastly, as an estimate for F&B, hotels appear to perform better in a high concentration of commercial and transport links functions, whilst Airbnb seems to perform better in highly residential areas proximate to the mainstream tourist attractions. Overall, this paper describes the use of the MGWR method in cases where localities is an important aspect of the spatial analysis process.
ISSN:2331-8422