Effect of Land Use and Road Network on Crime

Several studies exist that revealed decent association of crimes with distinct land uses and their transportation networks. However, the prevailing attempts are discrete in nature, i.e., lacked a systematic framework, and often focused on data from those land use where crime took place and ignored those land use data where crime did not take place. This study fills these gaps by proposing a new framework combining and synthesizing the notions of existing literature and evaluating it with the data from Uttara, a model town in Dhaka city, Bangladesh. The proposed framework commences by geocoding the crimes on land use GIS map, breaking down the map into different mesh sizes, associating crimes and no crimes with cells of those meshes and finding the optimum mesh size that can best explain the interrelationship among crime, land use and road network using Logistic Regression and their corresponding Receiver Operating Characteristic curves. Then it applies Support Vector Machine to identify various patterns of crimes in relation to land use and their distances from crime and no crime cells in terms of road network length. Several interesting connections were discovered, such as, built environment like street markets, highways, parks or open spaces are found prone to almost all types of crimes; murder and grievous crimes have poor relationship with land use; etc. which can provide insight into how to reduce crime in cities, how to allocate resources of law enforcement agencies, and/or how to design built environment which will act as natural deterrent to crime.

  • Supplemental Notes:
    • This paper was sponsored by TRB committee ADD30 Standing Committee on Transportation and Land Development.
  • Authors:
    • Sadeek, Soumik Nafis
    • Ahmed, Abu Jar Md Minhuz Uddin
    • Hossain, Moinul
    • Hanaoka, Shinya
  • Conference:
  • Date: 2018

Language

  • English

Media Info

  • Media Type: Digital/other
  • Features: Figures; Maps; References; Tables;
  • Pagination: 18p

Subject/Index Terms

Filing Info

  • Accession Number: 01663987
  • Record Type: Publication
  • Report/Paper Numbers: 18-06033
  • Files: TRIS, TRB, ATRI
  • Created Date: Mar 22 2018 12:03PM