Journal of Engineering and Engineering Technology - JEET |A FRAMEWORK OF DEEP LEARNING BASED TERRORIST CLASSIFICATION MODEL

Paper Details

Title :A FRAMEWORK OF DEEP LEARNING BASED TERRORIST CLASSIFICATION MODEL
Author(s) : Adewale Olumide S. (1), Jimoh Ibraheem T. (2), Makinde Ibrahim A. (3) and Adeleye Samuel A. (4)
DOI:
Abstract:

Terrorism has claimed many lives and destroyed many homes throughout the world. All levels of government have made significant investments in security, but there has not been much progress in releasing the terrified public from the grip of terrorism. Several agencies have deployed kinetic techniques such as tactical apprehension and killing of terrorists but the menace keeps increasing, hence, the exploration of soft computing which use intelligence gathering, community engagement and collaborative approach to combat the monster jerking the world without using forceful approach against the terrorists. The aim of the work is to propose a framework that predict the terrorism activities and offer useful information to the security agencies to prevent the risks of terrorist attack. The work proposes the use the Bidirectional Encoder Representation from Transformer for the world embedding that will be feed into the Attention-Based Bidirectional Long Short-Term Memory to analyze the tweets from Twitter, a social network service, to predict the activities of the terrorist group and classify the terrorist group that is responsible for each attack and provide necessary information for both the security agency and members of the society. The framework will be implemented with Python, Natural Language Processing and Deep learning packages such as Keras, TensorFlow, sklearn, NumPy, Pandas, Natural Language Tool Kits while the evaluation metrics such as accuracy, precision, recall, specificity, and F1 score will be used.

Pages

Contact us

Email us

Call Us