Efficient evaluation of continuous queries for XML data
Hidetaka Matsumura, Keishi Tajima · 2004
We developed efficient evaluation techniques for continuous queries, i.e. queries evaluated periodically, on XML data changing day by day. We assume queries are written in XPath, which is widely used today. Instead of completely reeval- uating queries, we compute the differences from the previous answers with minimal evaluation by using the information on changes. To achieve this, we maintain the information on deletion of which node in the data causes deletion of which node in the answers to which queries. In addition, when some node is added to the data, we can use the same information to omit the evaluation of part of queries.