Criminal behavior detection using the LSTM model and tweet dataset
Mayuri Diliprao Gaikwad, Abhimanyu Dutonde · 2024
The science of digital forensics is crucial in detecting and pursuing criminals since criminal behaviour depends more and more on digital technology. A model for behaviour detection is provided in this paper based on research into the most recent work in the field. when measured against models used in cutting-edge research. This study will help authorities working to prevent cybercrime and cybersecurity identify profiles with specific behaviours. The research focuses on identifying illegal behaviour from social media by evaluating the leading-edge work and suggesting an improved methodology. To get a better-trained model for this study, first collect the "Tweets Dataset for Detection of Cyber-Trolls" dataset from Kaggle. After that perform the preprocessing data for word cloud. the implementation work is complete on the python programming language and jupyer notebook. The results showed that the proposed LSTM model, achieved high accuracy. The suggested model outperformed with a 1.00% accuracy rate. This study will be very helpful to organization that deal with cybercrime and cyber security in identifying and eliminating profiles that exhibit specific behaviours.