Proximity search of XML data using ontology and XPath edit similarity

Abstract

XML data is explosively increasing, and a large amount of XML data, in which similar contents are described using different tag names and structures, have been emerging as a consequence. In such a situation, one cannot write a query against such XML data unless he/she knows the structure of the data. In this research, we propose a scheme to cope with this problem. Specifically, we expand XPath queries by replacing tag names with similar ones with the help of ontologies. In addition, we try to realize (structural) proximity matching of path expressions using edit similarity, which is a similarity measure based on edit distance. We also discuss application of SSJoin, which is an operator to support similarity joins in relational database systems, for speeding up the proposed scheme. We finally show the effectiveness of the proposed method by a series of experimentations. © Springer-Verlag Berlin Heidelberg 2007.