Prominent Binary-Feature (Permissions) Frequencies for Android Mobile Benign Apps and Malware Datasets

Published: 4 September 2020| Version 1 | DOI: 10.17632/ptd9fnsrtr.1
Contributor:
Gürol Canbek

Description

The following academic datasets provided the frequencies of prominent binary-features, namely Android application permission requests* for the negative and positive classes. They are used in analyzing the dataset dissimilarities via the method proposed by Gürol Canbek. Negative/Positive dataset pairs** and their references - DS0 by (Lindorfer et al., 2014). (ANDRUBIS) as it is called "Touchstone Dataset" in the related article - DS1 by (Aswini and Vinod, 2014). (Contagio) - DS2 by (Wang et al., 2014)*** - DS3 by (Yerima et al., 2014)*** - DS4 by (Jiang and Zhou, 2013). (Android Malware Genome Project, AMGP) - DS5 (Peng et al., 2012) Thanks to the authors providing the data. * https://developer.android.com/reference/android/Manifest.permission.html ** DS4 is a positive-class only dataset (it does not negative-class samples) *** DS2 and DS3 positive-class datasets contain DS4 (AMGP) samples.

Files

Categories

Cybersecurity, Mobile Computing, Machine Learning, Feature Selection, Mobile Device, Malware Mitigation, Similarity Measure, Data Quality, Multivariate Quantitative Analysis, Performance Evaluation, Classifier Evaluation, Kruskal-Wallis H Test

License