Mohammad Ghasemisharif
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Mohammad Ghasemisharif has a total of 9 co-authors. Here are the top co-authors by number of shared publications:
- Jason Polakis
University of Illinois Chicago, Chicago, USA
2 shared publications - Alberto Carboneri
University of Illinois Chicago, USA
1 shared publication - Daiping Liu
Palo Alto Networks, Santa Clara, CA, USA
1 shared publication - Janos Szurdi
Independent researcher, Budapest, Hungary
1 shared publication - Reethika Ramesh
Palo Alto Networks, Santa Clara, USA
1 shared publication - Ruian Duan
Palo Alto Networks Inc., Santa Clara, USA
1 shared publication
Latest Publications
- Open Access
- 4 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 1228-1245
Practical Real-time Detection of IPv4 Record Classical Domain Hijacking at Scale
One of the most dangerous cyberattacks is classical domain hijacking, where an attacker takes control of the administration of a domain's existing DNS records, modifying them, and sending visitors to malicious destinations. Domain hijacking can have a disastrous impact on both domain owners and their users, as criminals can direct all users to Man-in-the-Middle, phishing, and a wide array of other attacks. In recent years, researchers have analyzed and built detectors for various types of DNS hijacking attacks, including domain shadowing, exploitation of mistyped domains, and abuse of dangling records. However, only two studies on classical domain hijacking exist, failing to provide the community with a method to identify domain hijacking precisely and in time. This paper explores the feasibility of real-time classical domain hijacking detection for IPv4 records and characterizes its prevalence in the wild. We designed a framework that can process hundreds of millions of new records to identify a few dozen hijacking records or to analyze live DNS traffic from customers of a large cybersecurity company in around ten minutes. The framework leverages hundreds of terabytes of passive DNS logs and geolocation databases. It also utilizes active measurements such as WHOIS and web crawls to collect additional information about potential hijacking records. We evaluated the framework on our labeled datasets and additionally tested it in a large DNS security provider's network for over a year, finding multiple new domain hijacking attacks. Our work provides the first method for detecting domain hijacking with high precision, in real-time, and tested in a production environment.
- 4 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 1228-1245
Open AccessPractical Real-time Detection of IPv4 Record Classical Domain Hijacking at Scale
One of the most dangerous cyberattacks is classical domain hijacking, where an attacker takes control of the administration of a domain's existing DNS records, modifying them, and sending visitors to malicious destinations. Domain hijacking can have a disastrous impact on both domain owners and their users, as criminals can direct all users to Man-in-the-Middle, phishing, and a wide array of other attacks. In recent years, researchers have analyzed and built detectors for various types of DNS hijacking attacks, including domain shadowing, exploitation of mistyped domains, and abuse of dangling records. However, only two studies on classical domain hijacking exist, failing to provide the community with a method to identify domain hijacking precisely and in time. This paper explores the feasibility of real-time classical domain hijacking detection for IPv4 records and characterizes its prevalence in the wild. We designed a framework that can process hundreds of millions of new records to identify a few dozen hijacking records or to analyze live DNS traffic from customers of a large cybersecurity company in around ten minutes. The framework leverages hundreds of terabytes of passive DNS logs and geolocation databases. It also utilizes active measurements such as WHOIS and web crawls to collect additional information about potential hijacking records. We evaluated the framework on our labeled datasets and additionally tested it in a large DNS security provider's network for over a year, finding multiple new domain hijacking attacks. Our work provides the first method for detecting domain hijacking with high precision, in real-time, and tested in a production environment.
- Open Access
- 4 December 2023
- Published by Association for Computing Machinery (ACM)
- p. 44-55
When Push Comes to Shove: Empirical Analysis of Web Push Implementations in the Wild
Web push notifications are becoming an increasingly prevalent capability of modern web apps, intended to create a direct communication pipeline with users and increase user engagement. The seemingly straightforward functionality of push notifications obscures the complexities of the underlying design and implementation, which deviates from a near-universal practice in the web ecosystem: the ability to access an account (and the associated functionality) from practically any browser or device upon successful completion of the authentication process. Instead, push notifications create a communication endpoint for a specific browser instance. As a result, the challenges of deploying push notifications are further exacerbated due to the integration obstacles that arise from other aspects of web apps and user browsing behaviors (e.g., multi-device environments, account and session management). In this paper, we conduct an empirical analysis of push notification implementations in the wild, and identify common deployment pitfalls. We also demonstrate a series of attacks that target push notification functionality, including a novel subscription-sniffing attack, through a selection of use cases. To better understand current practices in push notifications implementations, we present a large-scale measurement of their deployment and also provide the first, to our knowledge, exploration and analysis of third-party service providers. Finally, we provide guidelines for developers and propose an approach for correctly handling push notifications in multi-browser, post-authentication settings.
- 4 December 2023
- Published by Association for Computing Machinery (ACM)
- p. 44-55
Open AccessWhen Push Comes to Shove: Empirical Analysis of Web Push Implementations in the Wild
Web push notifications are becoming an increasingly prevalent capability of modern web apps, intended to create a direct communication pipeline with users and increase user engagement. The seemingly straightforward functionality of push notifications obscures the complexities of the underlying design and implementation, which deviates from a near-universal practice in the web ecosystem: the ability to access an account (and the associated functionality) from practically any browser or device upon successful completion of the authentication process. Instead, push notifications create a communication endpoint for a specific browser instance. As a result, the challenges of deploying push notifications are further exacerbated due to the integration obstacles that arise from other aspects of web apps and user browsing behaviors (e.g., multi-device environments, account and session management). In this paper, we conduct an empirical analysis of push notification implementations in the wild, and identify common deployment pitfalls. We also demonstrate a series of attacks that target push notification functionality, including a novel subscription-sniffing attack, through a selection of use cases. To better understand current practices in push notifications implementations, we present a large-scale measurement of their deployment and also provide the first, to our knowledge, exploration and analysis of third-party service providers. Finally, we provide guidelines for developers and propose an approach for correctly handling push notifications in multi-browser, post-authentication settings.
- Open Access
- 15 November 2023
- Published by Association for Computing Machinery (ACM)
- p. 3475-3489
Read Between the Lines: Detecting Tracking JavaScript with Bytecode Classification
Browsers and extensions that aim to block online ads and tracking scripts predominantly rely on rules from filter lists for determining which resource requests must be blocked. These filter lists are often manually curated by a community of online users. However, due to the arms race between blockers and ad-supported websites, these rules must continuously get updated so as to adapt to novel bypassing techniques and modified requests, thus rendering the detection and rule-generation process cumbersome and reactive (which can result in major delays between propagation and detection). In this paper, we address the detection problem by proposing an automated pipeline that detects tracking and advertisement JavaScript resources with high accuracy, designed to incur minimal false positives and overhead. Our method models script detection as a text classification problem, where JavaScript resources are documents containing bytecode sequences. Since bytecode is directly obtained from the JavaScript interpreter, our technique is resilient against commonly used bypassing methods, such as URL randomization or code obfuscation. We experiment with both deep learning and traditional ML-based approaches for bytecode classification and show that our approach identifies ad/tracking scripts with 97.08% accuracy, significantly outperforming cutting-edge systems in terms of both precision and the level of required features. Our experimental analysis further highlights our system's capabilities, by demonstrating how it can augment filter lists by uncovering ad/tracking scripts that are currently unknown, as well as proactively detecting scripts that have been erroneously added by list curators.
- 15 November 2023
- Published by Association for Computing Machinery (ACM)
- p. 3475-3489
Open AccessRead Between the Lines: Detecting Tracking JavaScript with Bytecode Classification
Browsers and extensions that aim to block online ads and tracking scripts predominantly rely on rules from filter lists for determining which resource requests must be blocked. These filter lists are often manually curated by a community of online users. However, due to the arms race between blockers and ad-supported websites, these rules must continuously get updated so as to adapt to novel bypassing techniques and modified requests, thus rendering the detection and rule-generation process cumbersome and reactive (which can result in major delays between propagation and detection). In this paper, we address the detection problem by proposing an automated pipeline that detects tracking and advertisement JavaScript resources with high accuracy, designed to incur minimal false positives and overhead. Our method models script detection as a text classification problem, where JavaScript resources are documents containing bytecode sequences. Since bytecode is directly obtained from the JavaScript interpreter, our technique is resilient against commonly used bypassing methods, such as URL randomization or code obfuscation. We experiment with both deep learning and traditional ML-based approaches for bytecode classification and show that our approach identifies ad/tracking scripts with 97.08% accuracy, significantly outperforming cutting-edge systems in terms of both precision and the level of required features. Our experimental analysis further highlights our system's capabilities, by demonstrating how it can augment filter lists by uncovering ad/tracking scripts that are currently unknown, as well as proactively detecting scripts that have been erroneously added by list curators.