AU-PEMal-2025

The AU-PEMal-2025 dataset is created for research purposes, is a malware dataset that was created with proper care. A total of 15,500 benign and 15,500 latest malware samples were collected from VirusTotal, VirusShare, and MalwareBazar and organized into category-wise and family-wise folders. Each sample was run in a dynamic environment to analyze the functionality, and a detailed JavaScript Object Notation (JSON) file was generated against each sample. The sample includes four malware categories (RAT, Ransomware, Info Stealer, and Trojan Horse) and 26 malware families. Details of the malware categories and families as as below:- Sr. No. Category Families Total Samples 1 Trojan Horse 1. AgentTesla (337) 2. Emotet (345) 3. Guloader (444) 4. Qbot (320) 1446 2 Info Stealer 1. Formbook (473) 2. Raccoon (518) 3. RedLine (500) 4. Snake (527) 2018 3 Ransomware 1. Cerber (332) 2. DarkSide (335) 3. Dharma (505) 4. Gandcrab (382) 5. LockBit (438) 6. Maze (290) 7. Phobos (550) 8. Ragnar (133) 9. Revil (347) 10. Ryuk (449) 11. Shade (504) 12. WannaCry (496) 4761 4 RAT 1. Gh0st (446) 2. Glupteba (346) 3. NanoCore (520) 4. njRat (506) 5. Remcos (512) 6. Ursnif (309) 2639 5 Benign 1. Benign (10839) 10839 Selected features in the dataset are selected after a thorough analysis and by applying different techniques. The dataset is validated by training Machine Learning-based as well as Deep Learning-based models. AU-PEMal-2025-V1 contains original data in categorical form. AU-PEMal-2025-V2 contains numerical data with only target columns with categorical data.