Extracting global policies for efficient access control of XML documents

As documents containing sensitive information are exchanged over the Internet, access control of XML documents is becoming important. Access control policies can specify fine-grained rules to documents, but policies sometimes become redundant, as documents are restructured or combined during exchange. In this paper, we consider a new approach of optimizing access control policies, by extracting distribution information of given authorization values within XML data, The extracted information is called a global policy tree, and it can be utilized for minimizing the total size of policies as well as efficient query processing, We present a linear-time algorithm for minimizing policies utilizing global policy trees, and our evaluation results show significant improvement over existing work.