Volume 20 No 22 (2022)
 Download PDF
Exploring Convolutional Neural Networks for Malware Detection and Monitoring Social Media Activity
Priti Singh, Hari Om Sharan and C.S. Raghuvanshi
Social Media Networks (SMNs) serve as rapidly expanding communication platforms facilitating global connectivity for sharing messages, images, and videos. Given their integration with the cloud, SMNs pose heightened risks for malware and malicious activities. Users interacting with SMNs via Edge and Fog devices may inadvertently introduce malware through apps and APIs. Both one-to-one and one-to-many interactions on SMNs can be vectors for spreading malicious programs, impacting individuals and groups across various domains. While previous algorithms have targeted identifying malware-induced malicious actions, this paper introduces a Hybrid Machine Learning (HML) model leveraging advanced features of CNN and LSTM architectures for detection. The proposed HML integrates key features from both CNN and LSTM models, enhancing the CNN architecture with an additional folding layer and improving the pooling layer structure to K-max pooling. By dynamically adjusting the feature vector's length based on URL input size and CNN depth, the K-max pooling layer facilitates comprehensive feature extraction, while character and word embedding processes efficiently reduce data dimensionality. The HML model is implemented in Python using the KDD user profile dataset, yielding promising results with an accuracy of up to 95%, outperforming alternative models.
Malware Detection, Machine Learning, Social Media Network, Data Analytics, Malicious Activity.
Copyright © Neuroquantology

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Articles published in the Neuroquantology are available under Creative Commons Attribution Non-Commercial No Derivatives Licence (CC BY-NC-ND 4.0). Authors retain copyright in their work and grant IJECSE right of first publication under CC BY-NC-ND 4.0. Users have the right to read, download, copy, distribute, print, search, or link to the full texts of articles in this journal, and to use them for any other lawful purpose.