When Big Data Meets Thick Data: Deriving Evidence-Based Urban Design Principles for High-Density Neighbourhood Public Spaces

Keng Hua Chong (Singapore University of Technology and Design (SUTD))

Big Data refers to extremely large or complex data sets that could reveal patterns, connections and trends through computational analysis, which are useful in understanding human behaviour in urban environment. Yet what Big Data often lacks is the qualitative information to explain why certain things happen. While “Thick Data” refers to the qualitative aspects of human experience, gathered and analysed through ethnographic studies, which could reveal the underlying human emotions, stories, motivations, and models of operation. Works have begun recently in cross-analysing Big Data and Thick Data in consumer research. This paper attempts to apply such approach in urban design, to understand the urban dynamics particularly in the context of high-density neighbourhood public spaces, through visualising BOTH Big Data and Thick Data in an integrated platform, thereby deriving evidence-based urban design principles for selected spatial typologies.

The site selected for this research was Yuhua, a typical matured public housing neighbourhood in Singapore. Data were collected from multiple modes: Big Data through mobile application, place-based sensors, and social media: Thick Data through survey, interviews, participatory workshops, and field observations. A “Neighbourhood Public Space Framework” was developed to align and categorise the data sets into eight categories, each ranges between two distinct characters – People (sociable vs passive), Activity (lively vs restorative), Safety (user-friendly vs safety), Maintenance (functionality vs cleanliness), Convenience (amenities vs accessibility), Comfort (spaciousness vs environment), Aesthetic (landscape vs furniture), Identity (special to me vs special to us). Our data visualisation tool, Informed Design Platform (IDP), consolidates and visualises all the data according to these categories and variances, and enables interactive filtering of data based on different spatial, temporal, and demographic dimensions. The paper shows that through integrated, simultaneous visualisation of both Big Data and Thick Data, it is possible to derive contextual, evidence-based insights for responsive urban designs.