Back close

DeepAnti-PhishNet: Applying deep neural networks for phishing email detection CEN-AISecurity@IWSPA-2018

Publication Type : Conference Paper

Publisher : CEUR Workshop Proceedings

Source : CEUR Workshop Proceedings, CEUR-WS, Volume 2124, p.39-49 (2018)

Url : https://www.scopus.com/inward/record.uri?eid=2-s2.0-85050663201&partnerID=40&md5=d66eb1a441f5f25ae5aef2557b710c6f

Keywords : 10-fold cross-validation, Computer crime, Convolutional Neural Networks (CNN), Deep neural networks, Electronic mail, Feature representation, Fully-connected layers, Learning algorithms, Long short-term memory, Multi Layer Perceptron, Nonlinear activation functions, Recurrent neural network (RNN), Semantic similarity, Semantics

Campus : Coimbatore

School : School of Engineering

Center : Computational Engineering and Networking

Department : Electronics and Communication

Year : 2018

Abstract : Phishing represents a genuine risk to the Internet economy. Email has turned out to be a necessary verbal exchange tool in contemporary lifestyles. In recent days, email remains as the foremost generally utilized medium to dispatch phishing attacks. As a result, detection of phishing emails has been considered as an important task in the field of Cybersecurity. In this working note, we use word embedding and Neural Bag-of-ngrams with deep learning methods such as convolutional neural network (CNN), recurrent neural network (RNN), long short-term memory (LSTM) and multi-layer perceptron (MLP) to detect phishing email. Both word embedding and Neural Bag-of-ngrams facilitates to extract syntactic and semantic similarity of emails. Deep learning algorithms facilitate to extract the abstract and optimal feature representation and fully connected layer with non-linear activation function for classifcation. All the experiments are done on anti-phishing shared task corpus at IWSPA-AP 20181. All the models performed well during training phase. Moreover, word embedding with LSTM obtains 10-fold cross validation accuracy of 0.991 on sub task 12 and 0.971 on sub task 23. Based on the experimental results, we claim that word embedding with deep learning, specifcally LSTM is appropriate for the anti-phishing task. Copyright © by the paper's authors.

Cite this Research Publication : R. Vinayakumar, Ganesh, H. B. Barathi, M. Kumar, A., Dr. Soman K. P., and Poornachandran, P., “DeepAnti-PhishNet: Applying deep neural networks for phishing email detection CEN-AISecurity@IWSPA-2018”, in CEUR Workshop Proceedings, 2018, vol. 2124, pp. 39-49.

Admissions Apply Now