Content-aware malicious webpage detection using convolutional neural network

Malicious websites often install malware on user devices to gather user information or to disrupt device operations, violate user privacy, or adversely affect company interests. Many commercial tools are available to prevent malicious webpages from accessing devices; however, current versions of the...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Multimedia tools and applications Jg. 83; H. 3; S. 8145 - 8163
Hauptverfasser: Chang, Yen-Jen, Tsai, Kun-Lin, Jiang, Wei-Cheng, Liu, Meng-Kun
Format: Journal Article
Sprache:Englisch
Veröffentlicht: New York Springer US 01.01.2024
Springer Nature B.V
Schlagworte:
Accuracy
Algorithms
Artificial neural networks
Blacklisting
Computer Communication Networks
Computer engineering
Computer Science
Cybercrime
Data Structures and Information Theory
Deep learning
Internet
JavaScript
Keywords
Machine learning
Malware
Methods
Multimedia
Multimedia Information Systems
Neural networks
Social networks
Special Purpose and Application-Based Systems
Syntax
URLs
Visualization
Websites
Convolutional neural network (CNN)
Content awareness
Webpage contextual visualization
Malicious webpages
ISSN:1380-7501, 1573-7721
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Malicious websites often install malware on user devices to gather user information or to disrupt device operations, violate user privacy, or adversely affect company interests. Many commercial tools are available to prevent malicious webpages from accessing devices; however, current versions of these tools may become useless as soon as a new generation of malware is released. In this study, a content-aware malicious webpage detection (CAMD) method was developed; this CAMD method can verify whether a webpage is malicious by applying a novel webpage contextual visualization process, which retrieves the critical codes of webpages, transforms those codes into one-dimensional grayscale images, and applies convolutional neural networks to detect any malicious webpages. To verify the feasibility of proposed CAMD, 50000 normal and 50000 malicious webpages from the VirusTotal website were used. The results indicated that the proposed CAMD achieved an accuracy of > 98 % .
Bibliographie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1380-7501
1573-7721
DOI:10.1007/s11042-023-15559-8