Recently, keyword search has attracted a great deal of attention in XML database. It is hard to directly improve the relevancy of XML keyword search because lots of keyword-matched nodesmay not contribute to the results. To address this challenge, in this paper we design an adaptive XML keyword search approach, called XBridge, that can derive the semantics of a keyword query and generate a set of effective structured queries by analyzing the given keyword query and the schemas of XML data sources. To efficiently answer keyword query, we only need to evaluate the generated structured queries over the XML data sources with any existingXQuery search engine. In addition,we extend our approach to process top-k keyword search based on the execution plan to be proposed. The quality of the returned answers can be measured using the context of the keyword-matched nodes and the contents of the nodes together. The effectiveness and efficiency of XBridge is demonstrated with an experimental performance study on real XML data. © Springer-Verlag Berlin Heidelberg 2009.
|Title of host publication||Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)|
|Number of pages||12|
|Publication status||Published - 2009|
|Event||Advances in Data and Web Management, Joint International Conferences, APWeb/WAIM 2009 - Suzhou, China|
Duration: 2 Apr 2009 → 4 Apr 2009
|Conference||Advances in Data and Web Management, Joint International Conferences, APWeb/WAIM 2009|
|Period||2/04/09 → 4/04/09|