Enhancing IoT Security: Deep Learning Driven Web Phishing Attack Detection and Classification Model

The Internet of Things (IoT) is a technique that allows our day-to-day life objects to interconnect from the Internet and receive and send information for a meaningful intention. Nowadays, IoT has resulted in various revolts in nearly all areas of our societies. Phishing is the dominant threat to each internet user, in which invaders intend to falsely remove complex data from a system or user, with pretended websites, emails, and so on. Using the fast improvement in IoT gadgets, invaders are directing IoT devices like smart cars, security cameras, and so forth and performing phishing assaults to obtain control across such susceptible strategies for mischievous targets. This article focuses on the design of Al-Biruni earth radius optimization with Deep Learning for Web Phishing Attack Detection and Classification (BERDL-WPADC) technique. The presented BERDL-WPADC model majorly focuses on the identification and classification of web phishing attacks in the IoT environment. Primarily, the BERDL-WP ADC technique applies preprocesses the input data to convert it into a well-matched arrangement. Next, the classification of phishing web attacks occurs with a bi-directional LSTM (Bi-LSTM) classifier. For better the classification performance of the Bi-LSTM classifier, the parameter tuning process is performed through Al-Biruni earth radius optimization (BER) algorithm which supports attaining improved classification performance. The stimulated outcome study of the BERDL-WPADC model occurs and the outcomes are reviewed in terms of several aspects. The experimental investigation indicates the superiority of the BERDL-WPADC method across the existing methods.