SL handbook: objectives, scope and readership

The central objective of the SL handbook is to collect and catalogue the existing but often widely dispersed sources of information regarding spoken language resources within a single document. It is intended that the handbook will provide an essential reference work useful to a wide range of laboratories which are concerned with any aspect of speech technology. In addressing the production of the handbook, the project has kept in mind that the potential readership should include

• research workers and system developers who require convenient access to an organised body of specific reference material
• workers in other countries who require access to well-documented common practice in central Europe
• newcomers to the field who require introductory material, primarily research workers in related disciplines and students
• corporate end-users of speech technology, who need to specify, procure or integrate system components, and who require guidance related to system specification and assessment

The scope of the handbook fundamentally addresses the resources required for specifying, developing and evaluating speech technology components, including automatic speech recognition, speaker recognition and speech synthesis, which themselves are integrated to form interactive systems such as spoken dialogue systems. There is an emphasis upon the design and characterisation of speech corpora, the primary resource essential to both speech science and related technology developments, as well as upon assessment methodologies for the component technologies and integrated systems. It was specifically decided however not to attempt to address issues related directly to technology developments per se, such as methodologies and techniques for designing specific system components.