Enhancing Ad-blocking Systems for Better Web Browsing Security and Privacy
- Author
- Alrizah, Mshabab
- Published
- [University Park, Pennsylvania] : Pennsylvania State University, 2020.
- Physical Description
- 1 electronic document
- Additional Creators
- Zhu, Sencun
Access Online
- etda.libraries.psu.edu , Connect to this object online.
- Graduate Program
- Restrictions on Access
- Open Access.
- Summary
- Globally, hundreds of millions of Internet users worldwide are using ad-blocking systems. By blocking the ads and trackers on the web pages, Internet users can enhance security, privacy, performance, and data usage. Consequently, ad-blocking systems have drawn considerable attention from industrial and research communities. However, many aspects of such systems have not been studied in depth. The deficiency in understanding these aspects has become a common obstruction in optimizing ad-blocking systems. Therefore, the central theme of this dissertationis to measure the behaviors of different components of ad-blocking systems, and design techniques to optimize the efficiency and improve their security. Specifically, this dissertation makes the following contributions: (1) advancing the understanding of the ad-blocking community as well as the errors and pitfalls of the crowd sourcing process. We collected and analyzed several longitudinal datasets that covered the dynamic changes of a popular filter list named EasyList for nine years and the error reports submitted by the crowd in the same period. Our study yielded several significant findings regarding the characteristics of FP and FN errors and their causes. (2) boosting the efficiency of ad-blocking systems against evasion attacks. The robustness against evasion attacks is the primary metric to measure the effectiveness of ad-blocking systems. Therefore, we intensively analyzed evasion attacks from ad publishers against ad-blockers using data extracted from different sources. Our analysis revealed 15 types of attack methods, including 8 ways that have not been studied by the research community. (3) providing an analysis of ad-blocking systems against malicious attacks. We discovered and validated avulnerabilities of the crowdsourcing mechanism (with ethical experiments) and the software functionality. (4) optimizing the accuracy as well as the efficiency of ad-blocking systems while reinforcing the security by designing DeepSignature, a novel system that seamlessly integrate machine learning and crowdsourcing techniques.
- Other Subject(s)
- Genre(s)
- Dissertation Note
- Ph.D. Pennsylvania State University 2020.
- Reproduction Note
- Microfilm (positive). 1 reel ; 35 mm. (University Microfilms 29430894)
- Technical Details
- The full text of the dissertation is available as an Adobe Acrobat .pdf file ; Adobe Acrobat Reader required to view the file.
View MARC record | catkey: 30586047