Research And Implement Of Vulnerability Reuse Based On Multi-Source Data Fusion

In recent years,the Internet economy booming and computer technology developing rapidly,but network security situation has become increasingly severe.The emergence of security problems continues to alert practitioner.It is not difficult to find that the number of vulnerabilities increases each year,and the impact also expands by paying attention to the related platforms.Vulnerabilities can cause unpredictable harm to people's daily lives and even industrial production.Therefore,how to cope with endless vulnerabilities is a major challenge in the current network security industry.By analyzing the vulnerability data and security events,we suppose that the scope of the vulnerability is universal,and similar vulnerabilities frequently appear.Therefore,people can learn about the principle of known vulnerabilities,understand how to exploit it,and prevent most security issues.Aimed at these situations,this article proposes a protection proposal based on known vulnerabilities.Through the vulnerability database,we design a random forest-based vulnerability assessment prediction model to analyze the vulnerability hazards and implement a vulnerability scenario reuse platform to provide a vulnerability reproduction environment around massive vulnerabilities data with big data,machine learning and other technologies to build a network attack and defense knowledge map to provide support for complex cybersecurity issues.The main research contents of this article are as follows:Firstly,establish a vulnerability database.We use crawler to collect data from multiple databases at home and abroad,and extract vulnerability characteristics.We design and implement a multi-source vulnerability data fusion algorithm based on Simhash algorithm and calculate the similarity between vulnerabilities to achieve data fusion,so we provide the platform with more comprehensive vulnerability data information.Secondly,propose the vulnerability assessment and prediction model based on the random forest algorithm.At present,vulnerability assessments are mostly based on existing scoring models such as CVSS and DREAD,or analytic hierarchy process and attack graphs.Because of the lack of analysis of the original information of the vulnerability,our model starts from the vulnerability,based on the use of text word segmentation technology to extract the vulnerability characteristics,considers the relationship between vulnerability characteristics and hazard levels through random forest algorithm to predict the hazard level of the vulnerability,While further supplementing the vulnerability database information,it is also informative for the vulnerability reuse work.Finally,research on the vulnerability scenario reproduction platform.Due to the sensitivity of network attack technologies,and it is difficult to implement and verify offensive and defensive countermeasures in real networks.There are few approaches to study the principles of vulnerabilities and learn how to exploit them.To solve this problem,we use docker to provide an independent environment for each vulnerability to create a reuse scenario,and we also implement the basic functions and data storage based on the flask framework and postgresql to display the vulnerability information.