A new publicly available dataset WinAPI-4C-AdvMal with several categories of benign and malware software is collected and published in MIDAS. Windows API feature dataset for adversarial malware sample generation and obfuscation experiments (4 classes: office, network, security, malware). Data was obtained IAT (import address table). The dataset consists of 852 samples, from which 3591 unique features were extracted. Each feature is binary (0 – Win32 API not imported, 1 – Win32 API imported). Class 1 – Office/productivity tools. Class 2 – Internet/communication/networking tools Class 3 – Security related tools Class 4 – Malware. 

Olga Kurasova, Viktor Medvedev, Juozas Dautartas, Arnoldas Budžys, Haroldas Jomantas, & Juozapas Rokas Čypas (2025). WinAPI-4C-AdvMal: Windows API features for adversarial malware [dataset]. DOI: 10.18279/MIDAS.265677