KiloGrams: Very Large N-Grams for Malware Classification

TL;DR

This paper introduces a fast method for extracting very large n-grams (up to 1024) for malware classification, demonstrating their usefulness in creating interpretable features and industry-compatible signatures.

Contribution

It presents a novel, efficient approach to find top-k frequent large n-grams, enabling their use in malware detection and signature generation.

Findings

Large n-grams retain predictive power for malware classification.

The method is 60 times faster for small n and scalable to n ≥ 1024.

Large n-grams improve interpretability and signature creation.

Abstract

N-grams have been a common tool for information retrieval and machine learning applications for decades. In nearly all previous works, only a few values of $n$ are tested, with $n > 6$ being exceedingly rare. Larger values of $n$ are not tested due to computational burden or the fear of overfitting. In this work, we present a method to find the top-$k$ most frequent $n$-grams that is 60$\times$ faster for small $n$, and can tackle large $n\geq1024$. Despite the unprecedented size of $n$ considered, we show how these features still have predictive ability for malware classification tasks. More important, large $n$-grams provide benefits in producing features that are interpretable by malware analysis, and can be used to create general purpose signatures compatible with industry standard tools like Yara. Furthermore, the counts of common $n$-grams in a file may be added as features to publicly available human-engineered features that rival efficacy of professionally-developed features when used to train gradient-boosted decision tree models on the EMBER dataset.