Android Hybrid Apps Dataset

Published: 19 July 2021| Version 1 | DOI: 10.17632/bkjrvpg4br.1


This dataset has extracted features from Hybrid Apps available for deployment on the Android platform until recently. The data for this dataset has been culled out from various sources, including existing similar datasets and Google Play Store or its mirrors. The dataset is labelled to differentiate malicious and benign Hybrid Apps. Thus, it may conveniently be used for supervised learning. Nonetheless, the dataset has adequate attributes to support any unsupervised learning task as well. The dataset comprises 78,767 samples.


Steps to reproduce

The data was collected from multiple sources like Android Malware Dataset 2017 (CICAndMal2017)[1], Android Application Dataset for Malware Application [2], and Android Anti Malware Dataset [3]. Most attributes, which were not available in these Datasets, were extracted after downloading the APKs of these Apps from a mirror of Google Play Store, named 'APK Combo'[4].