ISCA International Workshop on Speech and Language Technology in Education (SLaTE 2011)
A database of non-native German productions was annotated
by three different groups: by experts using detailed, localised
labels as well as coarse, global labels, and by phoneticians
and na¨ýve subjects, using the same coarse global labels. For
the detailed annotation, segmental and supra-segmental labels
were given segment-based and word-based. The global annotation
consisted of a turn-based assessment of intelligibility, nonnative
accent, melody, and rhythm. Moreover, we use a large,
specialised prosodic feature vector for modelling native vs. nonnative
speech. We study relationships between detailed and
global labels, analyse the quality of expert and na¨ýve labellers,
and present an automatic system for predicting a speakers score
for the global labels.
Index Terms. non-native prosody, speech melody, rhythm, inter-labeller agreement, regression system, performance model, experts vs. naïve judges
Bibliographic reference. Hönig, Florian / Batliner, Anton / Nöth, Elmar (2011): "How many labellers revisited naïves, experts, and real experts", In SLaTE-2011, 137-140.