Modelling Airbnb locations and potential rents: A gravity-spatial interaction approach

Speaker: Dr Zara Shabrina (King's College London)

Abstract: Zara will talk about a specific case study on platform economy: Airbnb, a short-term rental platform for accommodation. She'll explore the possibility of predicting the locations of Airbnb and estimating the rental price for Airbnb listings in London's Lower Layer Super Output Area (LSOA). She does this by using the gravity-spatial interaction model based on her current work in collaboration with Robin Morphet at UCL.

Together, they examine how the rental price structure of Airbnb, compared with the Zoopla (British property website) rent data set, indicates the actual rent distribution of London housing. They implement two different types of gravity models: the singly (destination) constrained and the doubly constrained model.

The results indicate that the two approaches complement each other in terms of the information derived from both models. Based on their analysis, the estimated rents can explain for up to 64% of rental distribution in London. This implies that Airbnb has a similar distribution to the rental price distribution and is generally located in already highly-priced areas, which might further contribute to the increase in house prices in London.

This seminar will be preceded by tea, coffee and snacks in Room 6.05 from 12:30.

*If you are external to King's and would like to attend this event, please contact the event organiser directly.

Dr Zara Shabrina (King's College London) is a Lecturer in Geographic Data Science Education in the Department of Geography.

She received a BEng from the School of Planning, Bandung Institute of Technology Indonesia and a MUP from University of Southern California, USA, concentrating on preservation and design of the built environment.

She completed her PhD at UCL in 2019, after which she worked at the Bartlett Centre for Advanced Spatial Analysis (CASA) at UCL on determining measure of performance in smart and design based cities using varieties of quantitative and qualitative dimensions.