Access control strategy generation method and device based on machine learning

Abstract

The invention provides an access control strategy generation method based on machine learning, which comprises the following steps: acquiring an authorization log and historical authorization recordsof different access resources in an access log, and constructing an attribute tuple pair for the historical authorization records through the authorization log; training an attribute weight neural network through the attribute tuple pair; obtaining vector representation of the attribute according to the attribute weight neural network, and calculating correlation between access resources through the vector representation of the resource attribute; calculating the occurrence frequency of the attribute tuple pair, and generating a frequent item set when the occurrence frequency of the attributetuple pair exceeds a given threshold value; obtaining candidate rules according to the frequent item set; calculating the confidence coefficient and the support degree of the candidate rule accordingto the correlation between the access log and the access resource; and generating an access control strategy according to the confidence coefficient and the support degree. According to the invention,the similarity of access resources is utilized to help to increase the statistical strength and solve the problem of sparse authorization data.

Description

technical field [0001] The present application relates to the field of access control, in particular to a method and device for generating an access control policy based on machine learning. Background technique [0002] In the data access system, the access devices are connected to each other and are all connected to an open network environment, which brings severe challenges to data security and privacy protection. The current ABAC access control model formulates access policies based on user attributes, resource attributes, environment attributes, and operation attributes, and authorizes user access requests based on access policies. Because of the flexibility and ease of use of ABAC, it is widely used in access control, but the access control policy of ABAC needs to be formulated manually, and new access rules need to be continuously added during system operation, which is labor-intensive. The automated access control policy generation method can help administrators red...

AI Technical Summary

Problems solved by technology

Because ABAC's flexibility and ease of use are widely used in access control, but ABAC's access control policies need to be manually formulated, and new access rules need to be continuously added during system operation, which is labor-intensive.

The automated access control policy generation method can help administrators reduce the cost of manually formulating rules. Association rule mining technology can construct if-then rules. Using association rule mining methods to process historical authorization data in access logs can automatically formulate access policies. However, this method relies on a large number of authorization records of the target resource. If the number of authorization records of the target access resource is insufficient, it is difficult for the traditional access control policy generation method to automatically construct rules.

Images