Comparative Review of Different Techniques for Predictive Analytics in Crime Data Over Online Social Media

Social media platforms have swiftly become a vital platform for information sharing and communication. Its services are constantly being used by millions of people to connect with one another. Limiting and preventing crime, however, is one of the security agencies’ primary responsibilities when it comes to urban security. The development of enforcement strategies and the implementation of crime prevention and control depend heavily on crime prediction. The model can predict future instances of these crimes by using this strategy. Machine learning is currently widely utilized for predicting crime. However, in the age of big data, when individuals have access to an expanding amount of data, the capacity to recognize criminal patterns based on historical crime data is no longer effective. Models for deep learning and machine learning are contrasted using performance metrics. As a result, deep learning frequently surpasses machine learning when both of them are compared. A few datasets that are used for predicting crime statistics are studied along with their various classes. The analysis brings attention to the model's limitations as well as its possibilities for improvement.