مشخصات پژوهش

Because of quickly raising technology, internet has become a critical part of our daily lives. In response to the surge in internet apps, some recent efforts have been done for breaking into devices, networks, computer systems. Nonetheless, customers are supervising an increase in vulnerabilities and cyber-attack risks variety. Phishing refers to one of the feasible attacks. Phishing is broadly identified as a leading data breaches resource and the most prevalent deceitful cyber-attacks scam done by cybercriminals [1]. Phishing happens while an attacker persuades victim to do something that is not effective for system/ victim. Phishing is social engineering-based crime. New phishing has targeted organizations and made them bother because of the price including to contain productivity loses, malware, ransomware from phishing, credential compromise, besides loss of reputation in front of their users and competitors [2]. Phishers use different platforms, such as text messages, phone calls, email, forums, URLs, messaging apps for deriving info of a user. Their deceptive content sometimes imitates legitimate websites, enticing customers to communicate and divulge private info. The basic phishing target refers to identity theft/ financial gain resulting in businesses disruptions around the world [3]. Attackers are constantly evolving their ways to circumvent security mechanisms, leveraging advanced technologies like spear-phishing campaigns and employing psychological tactics to increase their success rates. This makes detecting phishing more challenging. This is feasible to protect such customers from phishing attacks applying different methods such as a strategy based on law, heuristic strategy, strategies based on URL, machine learning (ML). With deep learning (DL) and ML developments, this is currently feasible for obtaining high-accuracy mechanisms. For practical aims, such algorithms should appropriately group phishing and legitimate websites. Classification mechani