On Inference-Proof View Processing of XML Documents

Joachim Biskup,Lan Li

DOI: https://doi.org/10.1109/tdsc.2012.86

2013-01-01

Abstract:This work aims at treating the inference problem in XML documents that are assumed to represent potentially incomplete information. The inference problem consists in providing a control mechanism for enforcing inference-usability confinement of XML documents. More formally, an inference-proof view of an XML document is required to be both indistinguishable from the actual XML document to the clients under their inference capabilities, and to neither contain nor imply any confidential information. We present an algorithm for generating an inference-proof view by weakening the actual XML document, i.e., eliminating confidential information and other information that could be used to infer confidential information. In order to avoid inferences based on the schema of the XML documents, the DTD of the actual XML document is modified according to the weakening operations as well, such that the modified DTD conforms with the generated inference-proof view.