Churcher, GE, Souter, C and Atwell, ES (1996) Using a commercial speech recogniser within the domain of air traffic control. University of Leeds, School of Computing Research Report 1996.04
Abstract
We have taken an off-the-shelf, commercial continuous speech recogniser and conducted tests using three syntaxes for the domain of Air Traffic Control. The syntaxes are based on a corpus of transmissions between the ATC and pilots and reflect three differing levels of "linguistic" knowledge. The first represented the system where, in effect, there would be no syntax but a lexicon of all words in the corpus. The second took a partial look at syntactic information by using a key phrase spotting mechanism. The third represented the entire syntax of the corpus. Initial experiments show that key phrase spotting is insignificantly more accurate than no syntax at all, whilst use of a complete syntax can improve performance, to a point. The benefits of a discourse grammar are briefly discussed.
Metadata
Item Type: | Book |
---|---|
Authors/Creators: |
|
Copyright, Publisher and Additional Information: | Churcher, GE, Souter, C and Atwell, ES (c) 1996, University of Leeds. Reproduced with permission from the copyright holders. |
Keywords: | Airports |
Dates: |
|
Institution: | The University of Leeds |
Academic Units: | The University of Leeds > Faculty of Engineering & Physical Sciences (Leeds) > School of Computing (Leeds) > Artificial Intelligence & Biological Systems (Leeds) |
Depositing User: | Symplectic Publications |
Date Deposited: | 18 Dec 2014 11:38 |
Last Modified: | 16 Jan 2018 09:21 |
Published Version: | https://www.engineering.leeds.ac.uk/computing/rese... |
Status: | Published |
Publisher: | University of Leeds, School of Computing Research Report 1996.04 |
Open Archives Initiative ID (OAI ID): | oai:eprints.whiterose.ac.uk:81934 |