부산대학교 도서관

Phishing is a common tactic employed to deceive unsuspecting individuals into revealing their personal information through fraudulent websites. These phishing websites often steal passwords, usernames, and sensitive data related to online financial transactions. Phishers create websites that mimic the appearance and language of legitimate web pages to trick their targets. To combat the evolving techniques used in phishing attacks, it is crucial to implement antiphishing measures that can detect such attempts. Machine learning can play a significant role in stopping phishing attacks. This proposal presents a compilation of features and machine learningbased algorithms for identifying phishing attempts. Attackers frequently resort to phishing because it is easier to trick victims into clicking on seemingly legitimate yet malicious links, compared to circumventing computer security protocols. By replicating the official website's logo and other elements, the malicious link in the message body gives the impression of being affiliated with the legitimate website. This proposal explores the distinguishing characteristics of phishing domains, also known as scam domains, and emphasizes the importance of identifying them. Additionally, it discusses machine learning and natural language processing techniques that can be employed to achieve this goal.