A compressed accessibility map for XML

  • Authors:
  • Ting Yu;Divesh Srivastava;Laks V. S. Lakshmanan;H. V. Jagadish

  • Affiliations:
  • North Carolina State University;AT&T Labs--Research, NJ;University of British Columbia;University of Michigan, Ann Arbor

  • Venue:
  • ACM Transactions on Database Systems (TODS)
  • Year:
  • 2004

Quantified Score

Hi-index 0.00

Visualization

Abstract

XML is the undisputed standard for data representation and exchange. As companies transact business over the Internet, letting authorized customers directly access, and even modify, XML data offers many advantages in terms of cost, accuracy, and timeliness. Given the complex business relationships between companies, and the sensitive nature of information, access must be provided selectively, using sophisticated access control specifications. Using the specification directly to determine if a user has access to an XML data item can be extremely inefficient. The alternative of fully materializing, for each data item, the users authorized to access it can be space-inefficient. In this article, we introduce a compressed accessibility map (CAM) as a space- and time-efficient solution to the access control problem for XML data. A CAM compactly identifies the XML data items to which a user has access, by exploiting structural locality of accessibility in tree-structured data. We present a CAM lookup algorithm for determining if a user has access to a data item that takes time proportional to the product of the depth of the item in the XML data and logarithm of the CAM size. We develop an algorithm for building an optimal size CAM that takes time linear in the size of the XML data set. While optimality cannot be preserved incrementally under data item updates, we provide an algorithm for incrementally maintaining near-optimality. Finally, we experimentally demonstrate the effectiveness of the CAM for multiple users on a variety of real and synthetic data sets.