ABSTRACT
Speech applications should adopt language conventions that help people know what they should say next and avoid conversational patterns that violate standards of polite, cooperative behavior. A crucial factor in determining the success of a spoken application is whether there is a clear benefit to using speech technology. Speech is best used when it enables something such as conducting transactions that cannot otherwise be done over the telephone when a computer keyboard is not available, or for providing real-time automated translation services in a business setting for a user in conversation with another person speaking a different language. A digital record of the conversation is searchable, readable by the deaf and supports advanced business intelligence applications such as data mining. These are all examples of transcription tasks. Language translation applications build on the technological methods used in speech applications. These methods can be viewed as building blocks in the speech and language technologies domain.
