Clickjacking detection using Machine learning and Data Mining

Project ID: WUB/2022/P1/010

Project Duration: 2022 - 2023

Project Leader: MD Nazmus Sakib (Department of Computer Science & Engineering)

Project Members: Ashraf Kamal, Ayesha Siddika, Kazi Hassan Robin, Shamsun Nahar

Clickjacking attack is one of the emerging web-based attacks In clickjacking, transparent iframes that are put over website items are used to deceive the user into clicking. Without the user's awareness, this might result in undesirable operations. Although clickjacking is one of the main topics of debate, it is still unknown how and to what degree the attacker may employ this in order to motive user and get personal data. Therefore, the study have suggested a technique in this research that recognizes malicious URLs that are being utilized in clickjacking attacks. The webpage's dangerous links are categorized using the Self-Organizing Mapping (SOM) classification approach, and these links are then displayed on the webpage using HTML CSS properties. There is just one form of harmful link found on the webpage, and that is the phishing link. Thus, the phishing dataset is employed. Regarding performance measures, the Self-organizing mapping (SOM), the support vector machine learning model, and the SOM classification were compared. Compared to the SVM model and XGBoost, the SOM model requires shorter training time.