To address the current limitations of attribute-based access control(ABAC)policy generation methods,which are hin-dered by difficulties in extracting relationship information between attributes,as well as issues related to the quality and quantity of attributes,a hybrid policy generation method was proposed.The top-down approach was utilized to extract rich semantic access permission information words,eliminating the need to extract word relationships and thereby reducing the difficulty of the problem.Entity properties were improved based on semantic similarity optimization to reduce attribute quantity and improve quality.The deep forest model was improved to mine policies from the bottom up and improve the performance of access permis-sion decision-making under high-dimensional attributes.Experimental results show that the accuracy of the hybrid method in access decision-making can reach 98.11％,which is 2.05％higher than that of the direct single-generation method,while the policy model mining time is reduced by 21.53％.The proposed hybrid method is a more accurate and efficient ABAC policy gene-ration method.
维普
万方数据