A dataset of windows malware execution traces.

Area of Science:

• Cybersecurity
• Malware Analysis
• Data Science

Background:

• Malware poses a significant and evolving threat to digital security.
• Effective malware detection requires detailed execution trace data.
• Existing datasets often lack the necessary depth and diversity.

Purpose of the Study:

• To introduce a comprehensive dataset of Windows malware execution traces.
• To address the limitations of current publicly available malware datasets.
• To support advancements in malware detection and analysis.

Main Methods:

• Automated dynamic analysis of malware samples in a controlled virtualized environment (CAPEv2 Sandbox on Windows 10).
• Processing of raw sandbox reports using the MALVADA framework for filtering, structuring, labeling, and standardization.
• Generation of 31,844 standardized JSON execution trace files.

Main Results:

• A large-scale dataset of 31,844 detailed Windows malware execution traces.
• Each trace includes static metadata, dynamic behavioral information, and labeling fields.
• Standardized JSON format facilitates integration with data analysis and machine learning pipelines.

Conclusions:

• The presented dataset significantly enhances resources for malware research.
• It is suitable for developing and benchmarking malware detection methods, behavioral clustering, and automated labeling.
• The standardized structure promotes broader adoption and combination with other research data.