Android botnet detection dataset for machine learning

This dataset is generated from 6802 Android applications consisting of 1929 botnet apps from the ISCX botnet dataset and 4873 clean apps. The dataset has of 342 static features extracted from the apps and these features fall under 5 categories namely: API calls, Permissions, Extra (binary) files, Commands and Intents. The 1929 botnet apps consists of 14 different families. The dataset has been used to study different machine learning and deep learning techniques as described in the related papers.

本数据集源自6802个安卓(Android)应用程序，其中包含来自ISCX僵尸网络数据集的1929个僵尸网络应用，以及4873个良性应用。本数据集包含从上述应用中提取的342项静态特征，这些特征可划分为5大类：API调用、权限、额外（二进制）文件、命令与意图(Intents)。该1929个僵尸网络应用共涵盖14个不同的家族。本数据集已被用于开展各类机器学习与深度学习技术的相关研究，相关细节已在多篇关联学术论文中予以阐述。