Solving Cold Case in Police Department Using Machine Learning Algorithm
S Senthil Pandi, P Kanaga Shanmugam, R. Balaji · 2024
The unsolved cases in the records of police are often referred to as Cold Cases. The statistics of the crimes and attempt to crimes happening in India are eye opening. Close to 60,00,000 crime cases are registered in India every year and majority of these cases go unsolved and fall under the category of Cold Cases due to a number of factors like lack of evidence, lack of witness and lack of information. By using machine learning algorithms, the gap between the cases that are solved and the cases which are unsolved can be bridged. Taking in the important features of case files into consideration, the solved cases can be fed as an input to a model which is trained by clustering algorithms like K Means, DBSCAN etc. When the unsolved cases are fed to the trained model, the case files of the cluster can be suggested. These case files can then be used as references to solve the cases. The use case of this model can prove to be extremely beneficial in the context of solving age old cases where the records of those cases might be misplaced to lost or the evidences pertaining to those cases might not have been preserved for very long. While the project's scope is not restricted to solving crimes, if extended to the area of cybersecurity can have boundless applications as the threat level in the domain of cybersecurity is very high and if extrapolated, can help in protecting various cybercrimes from happening.