Loading...
Please wait, while we are loading the content...
Similar Documents
Join-Based Algorithms for Keyword Search in XML Databases
| Content Provider | Semantic Scholar |
|---|---|
| Author | Chen, Liang |
| Copyright Year | 2009 |
| Abstract | We consider the problem of keyword search in XML databases under the excluding lowest common ancestor (ELCA) semantics. Our analysis shows that ELCA semantics may lead to conflict with keyword proximity concept, and under such semantics, lower ELCAs are preferable because lower elements tend to be more specific. However, existing algorithms (stack-based and index-based) do not provide efficient support either for this the lower the better intuition or for general ranking functions, which is mainly due to the fact that generated results follow the document order. In this paper, we propose a join-based algorithm to compute complete ELCAs, which achieves complexity optimality for queries with various frequencies, as well as guarantees that lowest ELCAs are generated first. More importantly, we shed the new light on the connection between relational join and XML keyword search. Basically, many mature techniques in relational databases can be leveraged in this scenario to optimize query plan and improve execution efficiency. We further adopt the idea from top-K join in relational databases and propose a top-K algorithm for one type of ranking functions. Extensive experimental results demonstrate that the proposed algorithms outperform existing systems. |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://www.db.ucsd.edu:8080/root/pubsFileFolder/328.pdf |
| Alternate Webpage(s) | http://db.ucsd.edu/pubsFileFolder/328.pdf |
| Alternate Webpage(s) | http://www.db.ucsd.edu/pubsFileFolder/328.pdf |
| Alternate Webpage(s) | http://db.ucsd.edu/wp-content/uploads/pdfs/328.pdf |
| Alternate Webpage(s) | http://db.ucsd.edu/dblab-intranet/pubsFileFolder/328.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
