Research On Vulnerability Assessment And Associated Intelligence Recommendation Technology For Cloud Data Center

With the rapid development of cloud computing technology,cloud data centers have powerful storage and computing power,making it possible to store massive data and distribute computing.The migration of enterprises to the cloud,government affairs to the cloud,and personal data to the cloud has become an important feature of Internet services.Cloud data centers store a large amount of high-value data,and attacks on cloud data center infrastructure and tenant application vulnerabilities continue to increase.The security of cloud data centers is currently a major challenge.Carrying out asset-related vulnerability risk assessment and vulnerability mitigation measures is an important means of active security defense.However,the highly dynamic nature of cloud data center assets makes it difficult to assess vulnerability risks and obtain accurate security hardening strategies.Aiming at the above problems,a vulnerability assessment and related intelligence recommendation technology for cloud data center is proposed.The main research work of this paper is as follows:(1)In view of the fact that the current research on vulnerability assessment methods represented by CVSS lacks the analysis of the degree of correlation and impact on the assets themselves,and it is difficult to dynamically assess the vulnerabilities of cloud data centers,this paper is oriented to the real needs of cloud data centers,combined with the risk assessment.Based on theoretical knowledge and quantifying the importance of cloud data center assets,a vulnerability dynamic risk assessment method VRA-MDC is proposed for the analysis of cloud data center asset correlation impact degree,and finally the vulnerability scoring results from different perspectives are obtained.(2)In view of the current threat intelligence-driven active security hardening is difficult to obtain accurate threat intelligence support,based on the vulnerability risk assessment results,accurate associated threat intelligence is recommended for high-risk vulnerabilities in cloud data centers to achieve the effect of active defense.Aiming at the difficulty of obtaining semantic information from the short text information of vulnerabilities in traditional recommendation methods,this paper proposes a high-risk vulnerability related intelligence recommendation method CBL-R based on LDA topic model and BERT model in cloud data centers,which solves the characteristics of related intelligence recommendation process.The problem of sparsity and semantic ambiguity.The experimental results show that the recommendation performance consistency test result of the CBL-R model is 0.5124,which is better than the baseline method’s 0.2675,and has a better recommendation effect.