Data-driven Research on Street Environmental Qualities and Vitality Using GIS Mapping and Machine Learning, a Case Study of Ma On Shan, Hong Kong

Abstract

In a post-carbon framework, data-driven methods can be used to assess the environmental quality and sustainability of urban streetscape. Streets are an important part of people's daily lives and provide places for social interaction. Therefore, in this study, the relationship between street quality and street vibrancy is measured using the new town of Ma On Shan, Hong Kong as a study area. Firstly, machine learning was used to identify the physical features of streets through geographic information collection and streetscape image acquisition. Secondly, previous measurement algorithms are combined to calculate the greenness, walkability, safety, imageability, enclosure, and complexity of streets. Thirdly, secondary calculations and visualisations were carried out on a Geographic Information System (GIS) platform to observe the current distribution of street qualities. Finally, the relationship between street quality and vibrancy was analysed using SPSS statistical analysis software. The results show that walkability has a positive effect on street vitality, whereas safety and complexity have a negative effect on street vitality. This study demonstrates how the quantitative assessment of urban street environments can be used as a reference for building a green, lowcarbon, healthy, and walkable city.