Annotation

The annotation of a speech corpus refers to all symbolic information that is related to the speech signal, e.g. orthographic transcripts, phonemic transcripts, all kinds of segmentations. See chapter , pp. , for detailed discussion of annotation within the context of speech corpora.

Since in most cases some kind of annotation is an integral part of the speech corpus, you should define the contents of the desired annotation in the specification. Also it might be a good idea to define the procedures to achieve an annotation as well as the quality control of the annotation beforehand (see validation procedures, chapter and section ).