Detection of Phishing Attacks in PhiUSIIL Dataset using Deep Learning

In this modern era with a huge development in technologies and computation, the attackers are also emerging with new techniques or modified traditional techniques to break through the security of the systems and steal the confidential data for their own benefits. Phishing is a type of cyber attack, in this the attacker tricks the user by impersonating as an authentic & legitimate organization with the intentions of stealing the user’s credentials. From a long time and also recently phishing attack is constantly being in trend, many cyber-security organizations(such as OWASP, APWG, CISCO, etc) has released warnings and spreading awareness to stay alert from phishing. Hence this paper presents a phishing detection model. This proposed system combines the features of all pre-existing & similar technologies and adds a lot more to it. It is based on the branch of Machine Learning technology that is Deep Learning. It uses concepts of cybersecurity and analyze the dataset to detect if there is any potential phishing attack and alerts the user. This research compares the efficiency and accuracy of different models by training them for the recent dataset and concludes the best model that has the highest accuracy and efficiency for detection of phishing attacks. A user-friendly interface which can be used by individuals, security analysts, system administators and organizations. This model is a contribution to prevent one of the dangerous ongoing cyber attacks as it is using FCNN algorithm that is a Deep Learning algorithm trained using the most recent dataset available for phishing attacks. This model is capable to achieve higher accuracy and precision than general ML-based models.