Phishing detection using deep learning attention techniques

conference paper

English

In the modern world, electronic communication is defined as the most used technology for exchanging messages between users. The growing popularity of emails brings about considerable security risks and transforms them into an universal tool for spreading phishing content. Even though traditional techniques achieve high accuracy during spam filtering, they do not often catch up to the rapid growth and evolution of spam techniques. These approaches are affected by overfitting issues, may converge into a poor local minimum, are inefficient in high-dimensional data processing and have long-term maintainability problems. The main contribution of this paper is to develop and train advanced deep networks which use attention mechanisms for efficient phishing filtering and text understanding. Key aspects of the study lie in a detailed comparison of attention based machine learning methods, their specifics and accuracy during the application to the phishing problem. From a practical point of view, the paper is focused on email data corpus preprocessing. Deep learning attention based models, for instance the BERT and the XLNet, have been successfully implemented and compared using statistical metrics. Obtained results show indisputable advantages of deep attention techniques compared to the common approaches.

artificial intelligence; attention mechanism; deep learning; NLP; phishing filtering; text classification; transformers

SAFONOV, Y.

27. 4. 2021

Brno University of Technology, Faculty of Electrical Engineering and Communication

Brno

978-80-214-5943-4

Proceedings II of the 27th Student EEICT 2021 selected papers

1

131

135

5

https://www.fekt.vut.cz/conf/EEICT/archiv/sborniky/EEICT_2021_sbornik_2.pdf