Detecting Website Defacement Attacks using Web-page Text and Image Features

Recently, web attacks in general and defacement particular to websites applications have been considered one of major security threats many enterprises organizations who provide web-based services. A attack can result a critical effect the owner’s website, such as instant discontinuity website operations damage reputation, which turn may lead huge financial losses. number techniques, measures tools for monitoring detecting defacements researched, developed deployed practice. However, some techniques only work with static web-pages while others dynamic web-pages, but they require extensive computing resources. The other issues existing proposals are relatively low detection rate high false alarm because important elements embedded code images not processed. In order address these issues, this paper proposes combination model based on BiLSTM EfficientNet detection. proposed processes web-pages’ two components, including text content page screenshot images. effectively it produce accuracy well rate. Experimental results dataset over 96,000 confirm that outperforms models most measurements. model’s overall accuracy, F1-score positive 97.49%, 96.87% 1.49%, respectively.