Unified Pronunciation Services Project
The objective of the Unified Pronuciation Services Project is to create a single collaborative source of word pronunciations for text-to-speech synthesizers.This common shared resource can enhance the pronunciation facilities of text- to-speech synthesizers when words or sentences cannot be understood using the built-in pronunciation rules and exceptions in text-to-speech synthesizers.
Text-to-speech synthesizers may mis-pronounce acronyms, unfamiliar words, and mis-spelled words. Although some synthesizers allow users to fix the pronunciation for words they encounter, there is currently no way to easily share these pronunciations with others or to attach them to particular instances or Web sites. This project's collaborative web service will allow users to, on request or automatically through the text-to-speech synthesizer, look up unfamiliar words in the online table and receive the best pronunciation immediately from the web server. This project is complicated by language, regional dialects, text to speech engines, and preferences. Although synthesized speech may never perfectly match the user's own dialect, the goal is to allow people to visit unfamiliar pages or content and have the pronunciation be good enough for them to make sense of it.
Different ways to contribute
A peer collaborative model will be used to create and maintain the pronunciations. There are four types of contribution to this project:
- The development of the architecture and interface.
- The identification of existing (and freely available) databases of pronunciations.
- Once set up, ongoing peer contributions of pronunciations.
- Screening of contributions to remove SPAM and mischief.
Get involved
The discussion group is currently in formation stage. If you are interested in this project or have something to contribute include [PronunciationWG] in the subject line of an email to Contribution@RaisingTheFloor.net.
