NATICUSdroid: Android Malware Detection via Native and Custom Permissions

A 2021 dataset curated for the TabArena Tabular ML IID Study, intended for evaluating classification models on independent and identically distributed data. It originates from academic research on Android malware detection using native and custom permission features. The TabArena team removed a high proportion of duplicate rows present in the original data.

Use Cases

• Benchmarking binary classification models based on binary categorical permission features.
• Studying feature importance for malware detection based on Android permission patterns.
• Evaluating model performance on deduplicated, IID tabular data as part of a controlled study.

Strengths

• Released under a permissive CC BY 4.0 license.
• Curated to remove duplicate rows, which originally constituted 75% of the data.
• All features are binary categorical, which may simplify certain preprocessing and encoding steps.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count, file formats, and sample data are unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

https://doi.org/10.24432/C5FS64

Collection Method

Curated from original academic research for the TabArena Tabular ML IID Study.

Time Range

2021

Freshness

Dataset Year is 2021; last update date is unknown.

Geography

null