Zeng, Shi; (2024) Exploring Co-Location Patterns in London with Social Media and Spatial Data. Doctoral thesis (Ph.D), UCL (University College London).

Text Zeng_10201894_thesis_revised.pdf Access restricted to UCL open access staff until 1 July 2025. Download (7MB)

Abstract

This thesis explores the intersection of spatial co-location patterns and social phenomena through an innovative application of Twitter and other related data, addressing a research gap in existing spatial co-location studies which primarily focus on geographical phenomena alone. Spatial co-location pattern analysis is crucial for understanding spatial data and enhancing geographic context-awareness in applications. Traditionally, such studies have concentrated on identifying and analysing the spatial proximity of physical features or events to discern spatial interactions among different geographical phenomena. This research extends beyond mere geographical analysis to incorporate social phenomena, reflecting the real-world interconnection between geographic and social dynamics. By leveraging georeferenced Twitter data, this thesis aims to identify subsets of spatial features associated with social interactions and activities, offering a richer, more integrated view of how social and spatial elements interplay. The thesis proposes an innovative method to understand spatial co-location patterns from individual tweets at aggregated spatial levels. This includes finding spatial co-location mining techniques, analysing different categories (topics) of spatial co-location based on contextual information, and discovering unknown patterns that potentially influence the extent of current research. The research further explores co-location by enabling a comprehensive analysis using various forecasting techniques. Additional topics will be incorporated as covariates to evaluate the forecast models, and the results will reflect underlying co-location relationships. Finally, this thesis develops an approach for aggregated geographic co-location analysis (AGCA) by expanding previous examinations of geographic co-locations (GCL) from time series to geographic distributions and structures. This novel approach will provide a more comprehensive understanding of spatial co-location patterns and their dynamics. The findings of this thesis, whether conventional or unexpected, are expected to contribute significantly to the field from new perspectives by providing new insights into the relationship between spatial co-location patterns and social phenomena. The results will have practical implications for various domains, including urban planning, social science research, and location-based services, by offering a more comprehensive understanding of human behaviour and social interactions in the context of geographical space.

Download activity - last month

Export as XMLCSVJSON

Download activity - last 12 months

Export as CSVXMLJSON

Downloads by country - last 12 months

Export as JSONCSVXML

Archive Staff Only

View Item