A Survey on Malware Detection with Graph Representation Learning
Обзор методов обнаружения вредоносного программного обеспечения с использованием обучения представлений на графах
2024-05-21
SCID: 54.1/d8cf4jpd
Discuss with AI
control flow graphsfunction call graphsgraph neural networksgraph representation learningmalware detection
Figures from the paper
Abstract (AI)
Malware detection has become a major concern due to the increasing number and complexity of malware. Traditional detection methods based on signatures and heuristics are used for malware detection, but unfortunately, they suffer from poor generalization to unknown attacks and can be easily circumvented using obfuscation techniques. In recent years, Machine Learning (ML) and notably Deep Learning (DL) achieved impressive results in malware detection by learning useful representations from data and have become a solution preferred over traditional methods. Recently, the application of Graph Representation Learning (GRL) techniques on graph-structured data has demonstrated impressive capabilities in malware detection. This success benefits notably from the robust structure of graphs, which are challenging for attackers to alter, and their intrinsic explainability capabilities. In this survey, we provide an in-depth literature review to summarize and unify existing works under the common approaches and architectures. We notably demonstrate that Graph Neural Networks (GNNs) reach competitive results in learning robust embeddings from malware represented as expressive graph structures such as Function Call Graphs (FCGs) and Control Flow Graphs (CFGs). This study also discusses the robustness of GRL-based methods to adversarial attacks, contrasts their effectiveness with other ML/DL approaches, and outlines future research for practical deployment.
Key Findings
1
Graph Neural Networks achieve competitive performance in learning robust malware embeddings from expressive Function Call Graphs and Control Flow Graphs.
2
Graph Representation Learning has shown strong potential for malware detection by exploiting robust graph structures that are difficult for attackers to modify.
3
Graph-based malware detection offers intrinsic explainability and is reviewed alongside robustness to adversarial attacks and comparisons with other ML/DL methods.
4
The survey identifies future research directions needed to support practical deployment of Graph Representation Learning for malware detection.
5
Traditional signature- and heuristic-based malware detection methods generalize poorly to unknown attacks and are vulnerable to obfuscation.
Research Object
Malware represented as expressive graph structures, particularly Function Call Graphs (FCGs) and Control Flow Graphs (CFGs)
Research Subject
Graph Representation Learning (GRL) approaches for malware detection, including robust embedding performance, adversarial robustness, explainability, and comparative effectiveness against other ML/DL methods
Publication Details
Publication Date
2024-05-21
Journal
Publisher
ISSN
Cited by
98
Open access PDF
Access Type
Author Information
Download PDF
Subscribe to digest
References available in scid.ai17
A Unified Approach to Interpreting Model Predictions2017
Semi-Supervised Classification with Graph Convolutional Networks2016
Distributed Representations of Words and Phrases and their Compositionality2013
Neural Machine Translation by Jointly Learning to Align and Translate2014
Intriguing properties of neural networks2013
A Comprehensive Survey on Graph Neural Networks2020
The Graph Neural Network Model2008
Graph neural networks: A review of methods and applications2020
Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translation2016
Heterogeneous Graph Attention Network2019
Deep Learning on Graphs: A Survey2020
Graph Neural Networks: A Review of Methods and Applications2018
Graph Neural Networks in Recommender Systems: A Survey2022
Graph Neural Networks for Intrusion Detection: A Survey2023
Targeted Branching for the Maximum Independent Set Problem Using Graph Neural Networks2024
Qualitative Spatial Reasoning over Questions (Short Paper)2022
Detecting Functionality-Specific Vulnerabilities via Retrieving Individual Functionality-Equivalent APIs in Open-Source Repositories2025