Join-based Algorithm for XML Keyword Search

Jeff Chen (UCSD)

Keyword search in XML has attracted much attention in database community. It
allows users without prior knowledge of schema and query language (i.e. XQuery)
to exploit underlying data.  In this talk, I will first survey existing work on
XML keyword search, focusing on LCA-based (lowest common ancestor) semantics
and their algorithms.

Given the fuzzy property of the keyword query, top-K processing is desirable in
keyword search, because most of time users are only interested in very few
results with highest rankings. While the top-K issue is not well solved in XML
keyword search, there is a lot of work in relational databases, attacking the
problem from different dimensions. In the talk, I will take a close look at one
specific problem: top-K join.

Though top-K join from relational databases and keyword search in XML seem two
different domains, in the last part of the talk, I will discuss a join-based
algorithm for XML keyword search, shedding new light on the connection between
XML keyword search and relational join. In such scheme, top-K join algorithm
can be leveraged in XML scenario, providing an effective support for top-K
processing.