Loading...

Detection of Malicious Webpages

Arshad, Elham | 2016

548 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 48196 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Movaghar, Ali
  7. Abstract:
  8. Web applications and services have been developed and deployed with unprecedented speed, providing various important functionalities to the end user such as office applications, social networking, content sharing, education, and entertainment. Given its popularity and ubiquity, the Web also attracts the attention of malicious entities. Indeed, the Web and its global user community have observed various forms of attack in the past. Among these attacks, using the Web as a channel to distribute malware has become a prominent issue. This type of attack called drive by download attack. This issue has generated a great deal of attention from the security research community . Existing systems to identify malicious web pages fall into three main categories including honeypots, intra-page features-based and inter-page features-based. In this study, we described some previous research for each category and discussed the strength and weakness of each research. Finally, according to the characteristics and development of web-based attacks and the robustness of each category to evasion and obfuscation, the third category used to detect malicious webpages. In this thesis, the dataset was traffic of Computer department. Given the relationship between the pages, redirection chains and redirection graphs have been produced from dataset and every graph have been analyzed. Next, the features that distinguish the malicious redirection graphs from the benign ones have been extracted. Then by using machine learning techniques, recongnition pattern of malicious webpages has been created. The results show that the proposed method minimizes the limitations of the previous research and obtains more accurate detection
  9. Keywords:
  10. Identification ; Security ; Malwares ; Malicious Webpages ; Web-based Malware ; Malware Detection Signature

 Digital Object List

 Bookmark

No TOC