A DATA-DRIVEN APPROACH FOR INTERPRETING HUMAN PREFERENCE IN URBAN PUBLIC SPACES: A Case Study in London

Abstract

Sound, an often easily overlooked part of the urban composition, plays a crucial role in urban studies regarding how people perceive urban acoustic environments. Influenced by variable natural and artificial acoustic elements, the multi-level nature of urban soundscapes reflects multidimensional urban characteristics and thus demands thorough investigations. Unravelling the acoustic complexity and understanding its influence on public spaces' popularity will benefit the interpretation of urban soundscapes and help to build a more vibrant public life. In this paper, we utilize DepthMapX and QGIS to select specimen sites for further investigation. The behavioural analysis is performed through principle components analysis (PCA) and K-Means analysis based on data obtained within the city of London. This project seeks to develop digital strategies to investigate the relationship between urban public spaces and their soundscapes: to infer the correlation between people's emotions and the richness of urban environments. Through integrating machine-learning analytic approaches into visualizing how citizens perceive or experience the urban sound environment, our research aims to positively assist and inform urban practitioners in planning and designing more benign acoustic environments.