Location-Based Modelling And Analysis Of Threats By Using Text Mining
Sheshang Degadwala, Dhairya Vyas, Md Riyad Hossain, Abu Raihan Dider, Mohamed Nur Ali, Promise Kuri
Abstract
Location-Based Modelling and Analysis of Threats By Using Text Mining is process of analysis of threat and criminal behavior and help to the police to find the location of the criminal activities using some text mining techniques and machine learning algorithms. Now, in digital world, Computer crime such as identity threat, cybercrime, terrorist attack, rap, kidnaping, robbery has increase because of increase a digitalization has become a significant issue for police and law enforcement agencies Many identities threat occurs through social media and that's way crime are increases. This study helps to detect the all-criminal behavior-based activity and post on social media using different techniques and construct pattern sets based on domain knowledge from social media reports or newspapers and relevant literatures. In addition, our research include previous years crime activity is use for our model to train and that data is pre-process using tokenization for meaningful data. Some techniques such as feature extraction include emoticons and synonyms, while in classification Extra tree Classifier for analyses the criminal behavior from social media.