Anti-interference Approximate Role Mining RBAC System Construction

Role-based access control model(RBAC)is one of the hotspots of research on access control technology in the field of big data in recent years.When building an RBAC access control model system,the creation of roles is the core and the most critical part.Using humans to analyze the system and create roles requires a lot of manpower and requires very professional cyberspace security knowledge.However,using some existing accurate role mining algorithms will have problems such as generating a large role set and taking too long.The approximate role mining algorithm can effectively solve some of the above problems,but there will also be a series of problems such as affected by data noise,important roles or permissions lost,and the approximate coefficients are difficult to determine.The purpose of this study is to combine the data noise processing method to improve the approximate role mining algorithm to make it able to resist the impact of data noise,important roles or permission loss to a certain extent,and give recommendations for approximate coefficients in conjunction with the research.Aiming at the impact of data noise,the noise data detection theory based on Normal Degree is used to identify noise data.A noise-removing approximate role mining algorithm is designed,so that the approximate role mining algorithm can dynamically identify and correct the noise data during the role mining process.In view of the problem that important roles or permissions may be lost in the approximate mining algorithm,the characteristics of the mapping data and the defects of the mining algorithm are analyzed.Combining the mapping between the embedding degree of the user-permission mapping relationship and the role set,the important role mining algorithm is designed to To solve the problem of the loss of important roles or permissions in approximate role mining.Based on this algorithm,the approximate coefficient ? estimation method in the approximate role mining algorithm is given,which is used to reduce the interference caused by the common approximate role mining algorithm needing to set the ? value in advance without reference conditions.By solving the above problems,an anti-interference approximate role mining algorithm is proposed to improve the interference problem in approximate role mining to a certain extent,and an RBAC system architecture model based on the algorithm is given.Compared with the common approximate role mining algorithm,this algorithm can effectively reduce the impact of data noise,and can retain some important rolesand permissions with a certain degree of manual intervention so that it will not be lost by the algorithm.