Yujing Yang

Yujing Yang

Spatial Econometrics; Economic Geography; Regional Development; Transportation

Mapping Crime Exposure Risk in London

Yujing Yang

This analysis explores patterns of human mobility and exposure to urban crime risks in Greater London.

Analysis Overview

  • Objective: Detect and visualise spatial hotspots in Greater London where residents are most exposed to crime-related risks, leveraging location-based social network data and auxiliary urban datasets.
  • Approach:
    1. Risk Index Construction: Built a “threat index” using a Random Forest model
    2. Trajectory Analysis: Tracked user movement paths with smartphone data.

Key Findings

  • Fine-Grained Risk Mapping: Generated spatial risk maps at LSOA scale, revealing pronounced clusters of high exposure risk in Central and East London.
  • Resident-Filtered Analysis: By focusing on stable users (3+ months), the analysis reflects actual daily risks faced by London residents.

Visualization

A sample map below visualises the spatial distribution of threat indices and path trajectory across London: Path Trajectory Threat Index

Tools Used

  • Python: Data cleaning, machine learning (Random Forest), PCA
  • GIS: Spatial data processing,Trajectory Mapping
  • Open Data: Gowalla check-in dataset, ONS & Police UK crime data, OpenStreetMap POIs

Data Sources: Gowalla; Office for National Statistics; Police UK; London Datastore; OpenStreetMap