Chen, Tongxin; (2024) The impact of human mobility on urban crime: a geo-tagged big data-driven approach. Doctoral thesis (Ph.D), UCL (University College London).

Text Tongxin_Chen_PhD_UCL_final.pdf - Accepted Version Access restricted to UCL open access staff until 1 July 2025. Download (114MB)

Abstract

Urban crime is intricately tied to the dynamics of human mobility within the complex urban context. Understanding how human mobility shapes crime patterns provides valuable insights for both informing policy-making and urban development initiatives. This thesis delves into the impacts of human movement and visiting behaviours on crime occurrences across urban settings through a geo-tagged big data-driven methodological framework. Using extensive geo-tagged data sets sourced from mobile phone GPS trajectories in Greater London between 2020 and 2021, this research first utilised an integrated urban contextual approach to characterise human mobility patterns intertwined with different urban settings: it detected individual mobility trajectories from the raw GPS records, then generated and measured collective movement or visiting behaviour features connected to urban setting data. Subsequently, a set of machine learning techniques was employed to determine how population movement or visiting behaviour patterns explain variations in crime occurrences across urban neighbourhoods during different time periods over the pandemic. The results reveal nuanced insights into the interplay between human mobility dynamics and urban crime within the study area under various urban setting contexts. First, population routine dynamics exhibited a strong association with specific urban crime types, particularly property crimes and public order offences. Second, dynamic population routine activity factors showed a more pronounced influence on examined crimes than other static urban variables (e.g., land use and socioeconomic factors). Third, global impacts of population mobility (movement and visiting) behaviours on crime displayed fluctuations across different societal shifts, like pandemic periods. Last, further analyses revealed a diverse range of impacts of different population's mobility patterns on neighbourhood crimes, evidenced by variations in the influence of the population's activity related to urban place functions and neighbourhoods on the crimes at the localised level. In summary, this thesis used high-volume geo-tagged big data to demonstrate the profound influence of urban human mobility dynamics on crime patterns within different urban settings in Greater London, UK. The study reinforces the indispensable role of mobility big data in understanding the dynamic urban crime mechanisms and emphasises its potential to foster the development of a data-driven approach for crime intervention practice.

Download activity - last month

Export as JSONXMLCSV

Download activity - last 12 months

Export as XMLCSVJSON

Downloads by country - last 12 months

Export as JSONCSVXML

Archive Staff Only

View Item