Towards an inclusive system for the annotation of (dis)fluency in typical and atypical speech

ABSTRACT

This paper presents an operational annotation system for (dis)fluencies in typical and atypical speech, based on existing standard annotation schemes previously established in the literature. Grounded in a functional approach to (dis)fluency, we address some of the conceptual and technical limitations found in previous annotation models, and offer an integrated and inclusive system which is compatible with different multi-layered annotation software such as Praat or ELAN. Our aim is twofold: to create comparable annotated corpora both in typical and atypical speech, and to provide natural language processing and the health sector with applications for diagnostic and therapy in speech disorders.

KEYWORDS:

• Disfluency
• annotation
• typical speech
• atypical speech

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ The authors of this paper are ordered alphabetically. All authors contributed equally to all stages of the research project.

² Throughout this paper, we use the terms ‘typical’ and ‘atypical’ to refer to disfluencies that are common in all speakers vs. those that occur primarily in speakers presenting some speech or speech disorder, respectively. By extension, these terms are also used to characterise speech and speakers themselves, although this is by no means meant in a discriminatory way nor does it denies the right to be Neurodiverse.

³ Note that the terms ‘prolongation’ and ‘lenghtening’ are often interchangeable in the literature. For the sake of consistency, we are only using the latter in this paper.

⁴ ANR-18-CE36-0008 Le Bégaiement : la Neurologie, la Phonétique, l’Informatique pour son Diagnostic et sa Rééducation (BENEPHIDIRE, PI : Fabrice Hirsch).

⁵ A first version of this model was presented during the 2021 DiSS workshop (Didirková et al., 2021).

⁶ Krippendorff (2004) points out several limitations of Fleiss’ Kappa, such as its inability to account for individual preferences of annotators for particular categories. Despite its drawbacks, Kappa is widely used in corpus linguistics and is therefore useful to compare with previous proposals.

⁷ If κ < 0 : less than chance agreement; κ = 0.01–0.20: slight agreement; κ = 0.21–0.40: fair agreement; κ = 0.41–0.60: moderate agreement; κ = 0.61–0.80: substantial agreement; κ = 0.81–0.99: almost perfect agreement; κ = 1: perfect agreement..

Additional information

Funding

The work was supported by the Agence Nationale de la Recherche [ANR-18-CE36-0008]