Crime is an unlawful act punishable by a responsible authority in a country. In Sri Lanka, crime rate is moderate but there is no crime prediction system to minimize the criminal activities. Sometimes resource allocation of the defense authorities are not accurate because of this matter. Also, there has been an enormous increase in crime stats according to a research done by Australian data scientists. With the aim of securing the society from crimes, there is a need for advanced systems and new approaches for improving the crime analytics for protecting their communities. This review paper means to deliver the significance of new automated system for crime Identification and crime management in Sri Lanka.
Crimes are very dangerous and common social problem that faced by all countries in the world. These crime activities can affect human lives, economic of the country as well as the reputation. There are many types of crimes, but they can be divided into four types.
- Personal Crimes
- Property Crimes
- Inchoate Crimes
- Statutory Crimes
Personal crimes means offences against a person. This involves physical or mental harm to another person. These crimes include assault, battery, false imprisonment, kidnapping, homicide and rape or sexual assault.
Property crimes means offences against property. These crimes do not involve harm to another person. These crimes include theft, robbery, burglary, arson, forgery etc.
Inchoate crimes are crimes that were began, not completed. These crimes include attempt, solicitation and conspiracy.
Statutory crimes means violation of specific state or federal statute and can involve personal offence and property offence. So, minimize these kinds of crimes we need an accurate crime identification and management system.
In this project I paln to develop a system that can analyze the past crime data and find a pattern of crimes. Then the system can predict the crimes to be happen in next week, month or year. The result can be marked in the map for more information. Also, I hope to add the features of crime management and resource allocation of the police departments. I design this system for police departments, criminal investigation departments and Intelligence service Sri Lanka.
In this system I have five objectives.
- Analyze past crime data
- Find a pattern
- Mark the hotspot areas in the map
- Manage records
For analyzing phase, I plan to use 13 years of recorded past data of crimes in Sri Lanka. Back-end of the system will be in MySQL or MongoDB. So, the dataset will be stored in MySQL database or MongoDB database. Finding a pattern is the main and the hardest part of this system. For this phase I hope to use K-nearest neighbor algorithm or neural network. Frond-end of the system will be in java. From the results of the analysis, crime hotspot areas are marked in a map according to the corresponding district. Also, this system can manage all the records of past crimes, criminals and other records.
Main concern about this system is security. Because if someone who is willing to do unlawful things can get access to this system, he/she can get the information about all the ongoing processes of the defense authorities. To prevent that kind of scenarios, the system won’t be in web site or web page. This system can only install by the setup file those who have it. And only admins can login to the system. The design of the system is as follows.
These methods address one or more security properties and instantiate them into concrete security speciﬁcations/policies (either explicitly or implicitly) that are to be enforced at different phases in the lifecycle of the system.
In this paper it is a touching a new automated management system for crime hotspot area identification and crime prediction in the country. Further researchers should find ways to add more features to allocate the police officers by their duty number to corresponding crime scenes and record and store all the information through this automated system. With the availability of free tools and opensource projects these systems can be graphically stimulate and easily tested. The usage of these systems will prevent lot of errors in the current crime identification and crime management and make the world safe in the future.
Authors would like to acknowledge everyone who has supported in conducting this research including all the police officers of Kandy police station, crime investigation department and traffic department of Sri Lanka police, student of Management and Information technology batch 2013 – University of Kelaniya and Open University of Sri Lanka who helped with this research.