Design and Performance Evaluation of Crime Type and Crime Risk Score Estimation Technique for Fast and Efficient Response of Severe Crimes

摘要

In this paper, the machine learning-based estimation technique of crime types and crime risk score is introduced. For development of the estimation technique, the system platform of the introduced technique is also implemented. To predict the crime risk score, crime risk score equation is designed based on the crime severity and degree of damage, and numerical crime risk score is applied to each criminal case. For the crime types and crime risk score estimation, virtual police security data is considered. The data form is the same as the KICS data form. Using the proposed technique and implemented system platform, police can obtain efficient information in the first stage of criminal case reception.