A Predictive Model of Insider Threat Based on Bayesian Network

Hui Wang, Yunfeng Wang, Guangcan Yang

Abstract


At present, development of science and technology accelerates the society-informationization, many enterprises follow the trend of era to build internal network for convenient communication, but the increasing network security incidents cause a new understanding about the importance of internal network. The predictive model of insider threat based on Bayesian network is put forward in this paper. In the model, insider behaviors in the process of operation are considered as research objects, resource and intrusion evidence for operation sequence are seen as nodes, and then the network attack graph of Bayesian network is established. The concept of meta-operation, atomic attack and intrusion evidence are put forward in the graph. The node variable, its value and the conditional probability distribution of network attack graph are defined. Based on Bayesian network approximate inference, the improved likelihood weighted algorithm is presented to calculate the parameter and to quantify the insider threat. According to the simulation experiment data analysis, this approach is fast, simple and accurate, and plays an effective role in the process of insider threat prediction and evaluation.

Full Text:

PDF



International Journal of Online and Biomedical Engineering (iJOE) – eISSN: 2626-8493
Creative Commons License
Indexing:
Scopus logo Clarivate Analyatics ESCI logo IET Inspec logo DOAJ logo DBLP logo EBSCO logo Ulrich's logo Google Scholar logo MAS logo