Abstract
Predicting crime has become very important for the safety of civilians as there has been an unprecedented growth in crime related to social media. Police departments monitor social media platforms in a standard way to get information on crime. This project aims to facilitate law enforcement bodies in tracking crime data. We have proposed a system to classify the Twitter data into positive and negative sentiment tweets using deep learning. In the first part, we have explored several deep learning models to compare the result of sentiment analysis and implemented the model on the project. The Kaggle dataset has been used to train and test the model, which is subsequently larger than IEEE dataset. In the second part, we have used the computer visualization technique to shortlist the event from the challenge. We have developed several interactive graphs and dashboard systems to analyze the dataset in the project. For our research, we have taken IEEE’s vast challenge dataset to validate the system, and the results we have achieved are encouraging. The IEEE challenge is a fictitious scenario where the challenge encourages finding the whereabouts of the displaced employees using Twitter data.