Open Collections

Improving the portability of natural language interfaces through learning

University of British Columbia

One major concern with natural language interfaces is dealing with the problem of porting the system from one domain to another. Early approaches required that significant parts of the system be rewritten. Improvements have been made to the point that a system can be configured for a new database by a database expert who has no special knowledge of Computational Linguistics. Unfortunately, this has not yet lead to the widespread use of natural language interfaces. Proposed here is a natural language interface that can he configured by a user that has no knowledge of Computational Linguistics and no knowledge of databases ie - no knowledge of formal query languages, no knowledge of relational databases and no knowledge about how the data is organized in the database. Such a user merely needs to know some specific information about the domain that is modelled by the database. For an employee database, specific information is knowledge about a particular employee, eg - their name, address, and department.

Thesis/Dissertation

application/pdf

2008-09-15

For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use.

http://hdl.handle.net/2429/1985

Computer Science

University of British Columbia

UBCV

DSpace