Effective XML Representation for Spoken Language in Organisations

Abstract

Spoken Language can be used to provide insights into organisational processes, unfortunately transcription and coding stages are very time consuming and expensive. The concept of partial transcription and coding is proposed in which spoken language is indexed prior to any subsequent processing. The functional linguistic theory of texture is used to describe the effects of partial transcription on observational records. The standard used to encode transcript context and metadata is called CHAT, but a previous XML schema developed to implement it contains design assumptions that make it difficult to support partial transcription for example. This paper describes a more effective XML schema that overcomes many of these problems and is intended for use in applications that support the rapid development of spoken language deliverables.