Masquerade Detection Based on Temporal Convolutional Network

Masquerade detection has always been a vital detection capacity in intrusion detection, and shell command detection plays an important role in masquerade detection. Shell command detection is used to detect the system commands and judge whether the commands are from the masquerader or not to protect...

Full description

Saved in:
Bibliographic Details
Published in2022 IEEE 25th International Conference on Computer Supported Cooperative Work in Design (CSCWD) pp. 305 - 310
Main Authors Zhai, Haibin, Wang, Yong, Zou, Xueqiang, Wu, Yihan, Chen, Songyue, Wu, Hongwei, Zheng, Yanqin
Format Conference Proceeding
LanguageEnglish
Published IEEE 04.05.2022
Subjects
Computational modeling
Conferences
Deep learning
masquerade detection
Measurement
Neural networks
temporal convolutional network
Text categorization
text classification
Time series analysis
Online AccessGet full text

Cover

More Information
Summary:Masquerade detection has always been a vital detection capacity in intrusion detection, and shell command detection plays an important role in masquerade detection. Shell command detection is used to detect the system commands and judge whether the commands are from the masquerader or not to protect the safety of the whole system. However, it is challenging to classify the masquerader commands because of the difference between common text classification and shell command detection. The paper presents a machine learning model to masquerade detection using a Temporal Convolutional Network, a deep learning neural network for temporal anomaly detection. We believe masquerader commands are highly related to the time series. We prove that our model has a better effect on the SEA dataset than the deep neural network, convolutional neural network, and LSTM in various model metrics.
DOI:10.1109/CSCWD54268.2022.9776088