The relationship between residential burglaries and urban environments in Taipei City: a data mining approach

Although the overall crime rate in Taiwan has shown a declining trend in recent years, the proportion of burglaries remains high. Most studies regarding the prevention of burglaries proceed from the perspectives of population composition or criminal psychology, and place focus on the internal spatial planning of buildings without mentioning the association between case locations and urban environments. In order to effectively prevent crime and improve the quality of life, this study utilized cases provided by the Taipei City Government Open Platform (Data. Taipei) to confirm the surrounding environmental factors and building data of burglary cases for data mining application. The proposed method consists of two phases: clustering and association rule mining. In the clustering phase, the key substructures of environmental information are collected; then characteristics in each cluster are analyzed based on the association rule. The results of analysis showed that the first group (Group 1) in the classification should give priority to improving the visibility of idle space and streets, while the second group (Group 2) should improve the efficiency of personnel surveillance and solve the problem of crowds caused by business activities. The third group (Group 3), which featured narrow lanes and insufficient street lamps on chaotic streets, should engage in overall planning and design. Some studies have found that the environmental characteristics of burglary cases in Taipei City are different from the characteristics of crime-saddled urban areas defined by international scholars. Such differences deserve further study. The findings of this study can serve as an environmental design model for future urban development.