Identification of Hotspot of Rape Cases in NCT of Delhi: A Data Science Perspective

Abstract

Human history endorsed evidence of violence against women in different forms. Offense against women made an appearance in the early stages, and it is continuing. The present research based on an inferential methodology using scan statistics. It is one of the most extensively used statistical methods as one of the most popular emerging data science techniques in the study of events like crime. Moral degradation in human values, narrow-mindedness, growing intolerance, lack of value-based education, illiteracy, and unemployment are some of the critical factors, that is likely to be held responsible for offenders to perpetrate criminal activities that could lead to such a heinous and shameful crime event. The research is conducted based on the secondary data available on the data portal, gathered from different government sources. The geographical unit under the study is based on various crime zones adjusted to and embed into the revenue district of NCT of Delhi. The p-value obtained from the log-likelihood ratio for each crime district taken as the basis of identification and segregation of hotspot. The software used for computation is M.S. Excel, MS Solver, R-Studio, and SaTScan. The current work considered to be of great importance in the optimization of resources needed for crime control, monitoring, and surveillance to avoid such events in any form in future planning. The term optimization is used for maximal utilization of available scarce resources like installation of cameras, number of police deployments, distribution of various sophisticated policing equipment’s etc. to an extent of best possible measures to curb the menace effectively. Presently policing resources per person in NCT of Delhi is very much scarce and Delhi can be rated as very poor on that scale. For example, even in 2021 the estimated number of sanctioned strengths of Delhi Police is 83,762, means there are only 27 police sanctioned (not deployed) per ten thousand of the population. The actual number of deployed police per ten thousand populations may be further less. Similarly, we can cite an example of available CCTV cameras for surveillance estimated in the year 2021 is 1.32 lakh which comes to almost four CCTV for ten thousand person and even many of them are nonfunctional and defectives due to poor operational and maintenance reasons. Hence the problem of optimum utilization of resources is of prime importance. As such the relevance of the present work is aiming at societal interest and also enables our stressed policing bodies to execute an effective planning, is likely to be appreciated.