Conference Paper/Proceeding/Abstract 1634 views
Acquiring Pronunciation Data for a Placenames Lexicon in a Less-Resourced Language
Briony Williams,
Rhys Jones 
Proceedings of the Sixth International Conference on Language Resources and Evaluation (LREC'08)
Swansea University Author:
Rhys Jones
Abstract
We describe a procedure for generating pronunciations for a place-name dictionary for Welsh. For many such less-resourced languages, phoneticians with expertise in the language do not exist, so direct editing of pronunciations cannot occur. We describe a method where, if a native speaker and a text-...
| Published in: | Proceedings of the Sixth International Conference on Language Resources and Evaluation (LREC'08) |
|---|---|
| Published: |
Marrakech, Morocco
European Language Resources Association (ELRA)
2008
|
| Online Access: |
http://www.lrec-conf.org/proceedings/lrec2008/pdf/55_paper.pdf |
| URI: | https://cronfa.swan.ac.uk/Record/cronfa11776 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Abstract: |
We describe a procedure for generating pronunciations for a place-name dictionary for Welsh. For many such less-resourced languages, phoneticians with expertise in the language do not exist, so direct editing of pronunciations cannot occur. We describe a method where, if a native speaker and a text-to-speech synthesiser is available, pronunciations can be edited without detailed knowledge of the language's phonology. Our key development is the use of 're-spelling' to indicate, in a linguistically-naïve manner, the required pronunciation. These drive a set of specially-adapted letter-to-sound rules, thus generating the desired pronunciation. A speech synthesiser gives audio feedback to the native speaker and verifies that the pronunciation is correct, and a graphical user interface links the database backend, the speech synthesiser and the native speaker editor. This method has been used to generate a pronunciation database for Welsh placenames. |
|---|---|
| Keywords: |
Lexicon, lexical database, Speech synthesis, Text-to-speech systems, Tools, systems, applications |
| College: |
Faculty of Humanities and Social Sciences |