Speech Application Design and Development

The use of speech is expected to make the human-computer interaction more natural and efficient than it has been so far. Many successful speech applications have been constructed, but speech is still far from being the most common input and output modalities. In order to successfully construct speech applications, system designers and implementers must know the nature of speech, the characteristics of the underlying technology and interaction techniques and models that are suitable for speech-based human-computer communication. In this chapter the most important issues of speech applications from this viewpoint are presented. synthesis are discussed from the application development perspective. Technical presentation is omitted and the focus is on those issues which have an impact on speech-based human-computer interaction. For a technical introduction, a guidebook such as [Huang et al., 2001] is recommended. After the technology overview, an introduction to speech applications is presented. This includes conventional speech applications, such as telephony systems, and new application areas, such as multimodal and pervasive applications. Applications are presented from the viewpoint of application development, including interface issues and the underlying system architecture. The last two sections present application design principles and evaluation methods. Issues such as the nature of speech and speech-based communication, evaluation metrics and empirical experiments are also presented. Although speech is the most natural form of communication between humans, it has several limitations when used in interaction between humans and computers. Some of its advantages which are present in human-human communication cannot be used in human-computer interaction. There are multiple reasons for this, both technical shortcomings and those which are related to the nature of speech as a communication modality. In many cases, communication is not based on the best possible solutions, but instead the technology limits choices and even dictates the design [Mane et al., 1996]. In order to successfully use speech in the interface, we must know the state of the art and reality of speech technologies, as well as what kind of interface element speech is. First, it is important to realize that speech-based human-computer interaction is different from written language and spoken communication between humans. Secondly, speech technology has many technical limitations, and it is not realistic to believe that these will disappear in the near future. In this section those issues of speech technology which affect spoken interaction are presented. Speech can be structured in many different ways depending on its use. One useful …