Detecting semantically correct changes to relevant unordered hidden web data

Abstract

Current proposals for XML change detection use structural constraints to detect the changes and they ignore semantic constraints. Consequently, they may produce semantically incorrect changes. In this paper, we argue that the semantics of data is important for change detection. We present a semantic-conscious change detection technique for the hidden web data. In our approach we transform the unordered hidden web query results to XML format and then detect the changes between two versions of XML representation of the hidden web data by extending X-Diff, a published unordered XML change detection algorithm. By taking advantage of the semantics, we experimentally demonstrate that our change detection approach runs up to 7 times faster than X-Diff on real life hidden web data and always detect changes that are semantically more correct than those detected by existing proposals. © Springer-Verlag Berlin Heidelberg 2005.