Mapping the property crime spatial pattern in Selangor using social media data mining and GIS

The previous studies had shown that the information extracted from social media can be utilised in locating offenders, establishing probable cause for warrants, and identifying potential witnesses. By integrating social media data mining with Geographic Information System (GIS) techniques, it is possible to map property crime hotspots based on user-generated content. This approach can provide valuable insights to complement officially reported crime data. Through GIS-based spatial analysis, patterns and distributions of crime can be identified, enabling the detection of highcrime areas or hotspots. The ability of hotspot analysis to present crime concentration across a geographical landscape makes it a powerful and practical tool for law enforcement and urban planning. From this study, the total property crime geocoded data that managed to collect is 488 cases consisting of the snatch, burglary, theft, and car theft crime in Selangor. Each crime case has been analysed for its spatial pattern and distribution where for both snatch and burglary crime exhibits clustering while theft crime gives pattern and distribution of random across the study area. The spatial pattern of each crime show that the southern part of Selangor has more crime cases as compared to northern part of Selangor. The red zones show the areas with a very high value of z score indicate significant spatial clustering The findings would be beneficial to the relevant authorities regarding the underreporting of crime cases.