Malware-Detection-Using-a-Single-Bidirectional-Graph-Embedding

overview

We are interested in the detection and classification of malware based on a new graph coding framework. The single-bi-directional graph coding framework is based on a Keras implementation.

Datasets

The dataset is too large to upload here. We provide sample data as well as the source of the data. We used the data pre-processing approach in the paper to process the data from the original API call sequence of the sample software and obtain the "sample_data.csv".
The first dataset's data is from the Alibaba-Security-Algorithm-Challenge, and the related web site is: https://tianchi.aliyun.com/competition/entrance/231694/information
The second dataset's data is from the work of “Deep learning based Sequential model for malware analysis using Windows exe API Calls”, and the related web site is: https://github.com/ocatak/malware_api_class

Usage:

1.Graph_Embedding

python Graph_Embedding.py
See code(Graph_Embedding.ipynb)for details

2.Predict

python predict.py <api_number> <seq_length>
api_number:Interer,Number of API categories
seq_length:Interer,Length of model input sequence
See code(Predict.ipynb)for details