Lecturer: Doc Li Guoliang
Lecture Time: Apr 22 2010, 10:00am-11:30am
Lecture Place:
Abstract:
Traditional query processing approaches on relational and XML databases are constrained by the query constructs imposed by query languages such as SQL and XQuery. First, the query languages themselves can be difficult to comprehend for non-database users. Second, these query languages require queries to be posed against the underlying, sometimes complex, database schemas. Third, although some applications can be provided on top of databases to facilitate easy access of the underlying data, it is rather difficult to devise a versatile application which can adapt to any underlying database. Fortunately, keyword search is proposed to provide an alternative means of querying relational databases, which is simple and familiar to most Internet users. One important advantage of keyword search is that the users can search information without having to use a complex query language such as SQL or XQuery, or to understand the underlying data structures. This talk will discuss efficient algorithms, effective indexing structures, and ranking mechanisms to support efficient keyword search over structured data.
