Tools
Tools
Salian, SaiPrasad Sadashiv (2024) Enhancing Phishing Email Detection with Sentiment Analysis: A Hybrid Approach. Masters thesis, Dublin, National College of Ireland.
Preview |
PDF (Master of Science)
Download (1MB) | Preview |
Preview |
PDF (Configuration Manual)
Download (944kB) | Preview |
Abstract
In this digital era, people are surrounded by technology and the internet usage has skyrocketed. This means that there’s more data available on the internet about a particular person than they can dream of. Cybercriminals use these data to launch attacks via emails and try to steal their sensitive information. Phishing is a very common type of attack used by these criminals and they try to attack large organizations in order to obtain ransom from them or to deal with their information. These attacks happening inside an organization through an organizational email has the potential to lead losses in billions. To detect such malicious attempt, this paper proposes a phishing email detection system that can analyse the sentiment behind the email and improve the level of accuracy. We use the DistilBERT model for sentiment analysis and then feed the sentiment aware embeddings to a SVM model for further classification. This proposed study provides steps on how to create a system that can not only check the regual heuristic features and key filters but also the sentiment behind the email which will help improve the overall security.
Actions (login required)
 |
View Item |