Skip to Main content Skip to Navigation
Conference papers

Investigation par méthodes d'apprentissage des spécificités langagières propres aux personnes avec schizophrénie

Abstract : We present experiments to automatically identify patients with symptoms of schizophrenia in controlled conversations between patients and sychotherapists. We merge the speech turns of each interlocutor and train basic classifiers with these data, using lexical, morphological and syntactic features. This study is the first of its kind in French and obtains results comparable to the English state-of-the-art. Our first experiments highlight that the speech of patient with schizophrenia differs from control one : the best model obtains an accuracy of 93.66%. However, richer descriptions will be needed to produce an applicable and reliable model.
Complete list of metadata

Cited literature [36 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-02784752
Contributor : Sylvain Pogodalla Connect in order to contact the contributor
Submitted on : Tuesday, June 23, 2020 - 11:53:24 AM
Last modification on : Monday, July 11, 2022 - 10:33:43 AM

File

123.pdf
Publisher files allowed on an open archive

Licence


Distributed under a Creative Commons Attribution 4.0 International License

Identifiers

  • HAL Id : hal-02784752, version 4

Citation

Maxime Amblard, Chuyuan Li, Chloé Braud, Caroline Demily, Nicolas Franck, et al.. Investigation par méthodes d'apprentissage des spécificités langagières propres aux personnes avec schizophrénie. 27èmes Journées Traitement Automatique des Langues Naturelles (TALN 2020), Jun 2020, Nancy, France. pp.12-26. ⟨hal-02784752v4⟩

Share

Metrics

Record views

293

Files downloads

178