Efficient top-k algorithm for eXtensible Markup Language keyword search

The ability to compute top-k matches to eXtensible Markup Language (XML) queries is gaining importance owing to the increasing of large XML repositories. Current work on top-k match to XML queries mainly focuses on employing XPath, XQuery or NEXI as the query language, whereas little work has concerned on top-k match to XML keyword search. In this study, the authors propose a novel two-layer-based index construction and associated algorithm for efficiently computing top-k results for XML keyword search. Our core contribution, the two-layer-based inverted Index and associated algorithm for XML keyword search take both score-sorted-sequence and Dewey ID-sorted-sequence into consideration, and thus gain performance benefits during querying process. The authors have conducted expensive experiments and our experimental results show efficiency advantages compared with existing approaches.