Looking at the bag is not enough to find the bomb: an evasion of structural methods for malicious PDF files detection

By Davide Maiorca, Igino Corona, Giorgio Giacinto. 2013

152 citations recorded by the source corpus.

Subfields: Computer Networks and Communications, Information Systems, Privacy & Security, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Networks and Communications

In Information Systems

In Privacy & Security

In Signal Processing

Citations in the corpus

Is Feature Selection Secure against Training Data Poisoning? (2015, methodology): “For instance, detection of malware in PDF files relies on the analysis of the PDF logical structure, which consists of a large set of different kinds of objects and metadata, yielding a high-dimensional data representation (Smutz & Stavrou, 2012; Maiorca et al., 2012; 2013; Šrndić & Laskov, 2013).”
Practical Evasion of a Learning-Based Classifier: A Case Study (2014, background): “An inverse attack, performed by injecting malicious content into a benign PDF file, was described in [33] and demonstrated to be effective against PDFRATE in a small-scale experiment.”
Automatically Evading Classifiers: A Case Study on PDF Malware Classifiers (2016, neutral): “For example, previous studies may mistakenly assume the attackers can only insert new contents because removing existing contents would easily corrupt maliciousness [4, 20, 28].”
Adversarial Feature Selection Against Evasion Attacks (2016, methodology): “of malware samples, by manipulating their code [17]), and considering such manipulations as the only ones available to the attacker to find evasion points.”
One-and-a-Half-Class Multiple Classifier Systems for Secure Learning Against Evasion Attacks at Test Time (2015, methodology): “In this experiment, we use the feature representation adopted in [22,8], in which each feature represents the number of occurrences of a given keyword in a PDF file.”
Intriguing Properties of Adversarial ML Attacks in the Problem Space (2019, background): “Existing work investigated problem-space attacks on text [3, 43], malicious PDFs [12, 22, 41, 45, 46, 74], Android malware [23, 75], Windows malware [38, 60], NIDS [6, 7, 20, 28], ICS [76], source code attribution [58], malicious Javascript [27], and eyeglass frames [62].”
A Survey on Security Threats and Defensive Techniques of Machine Learning: A Data Driven View (2018, background): “Similar to the above attack, the impersonate attack prefers to imitate data samples from victims, particularly in application scenarios of image recognition [59], malware detection [60], [61], intrusion detection [62] based on machine learning.”
Feature Conservation in Adversarial Classifier Evasion: A Case Study (2017, background): “The ﬁrst involves attacks in problem space: systematic approaches for designing evasions through modifying actual malicious instances [30], [37], [21].”
PlatPal: Detecting Malicious Documents with Platform Diversity (2017, background): “Reverse mimicry [35] attack, on the contrary, attempts to embed malicious content into a benign PDF by taking care to modify it as little as possible.”
When a Tree Falls: Using Diversity in Ensemble Classifiers to Identify Evasion in Malware Detectors (2016, methodology): “To evaluate our technique, we applied mutual agreement analysis to two well-studied malware detection systems: PDFrate [40] and Drebin [4].”
Hardening Classifiers against Evasion: the Good, the Bad, and the Ugly (2017, background): “The first involves attacks in problem space, which construct evasions through modifying actual malicious instances [8], [31], [37], [21].”
Man vs. Machine: Practical Adversarial Detection of Malicious Crowdsourcing Workers (2014, background): “Several studies have examined attacks on specific MLbased applications, from email spam detection [12] to network intuition detection [37, 40, 44] to malicious (PDF) file classification [5, 25, 41] and malware detection [21].”
Pattern Recognition Systems under Attack: Design Issues and Research Challenges (2014, background): “4.1, and how to re-engineer their design cycle, using a proactive approach.”
Towards Adversarial Malware Detection (2018, background): “[63] 2013 Heur.”
Digital Investigation of PDF Files: Unveiling Traces of Embedded Malware (2017, background): “In terms of the attacker’s knowledge, mimicry [4], [8] and reverse mimicry [12] attacks do not exploit any specific knowledge of the attacked system.”
Cracking Classifiers for Evasion: A Case Study on the Google's Phishing Pages Filter (2016, methodology): “In [41], a simple but effective attack methodology called reverse mimicry is designed to evade structural PDF malware detection systems.”
CryptoLock (and Drop It): Stopping Ransomware Attacks on User Data (2016, background): “Furthermore, recent research has shown that evading signature detection is possible with relative ease when the malware signatures used are too rigid [9], [24], [32], [33].”