Employing large language models for emotion detection in psychotherapy transcripts
Even though the classification results were only moderate, our model achieved a good performance especially for prediction of symptom severity. The results confirm the role of negative emotions in the prediction of symptom severity, while they also highlight the role of positive emotions in fostering a good alliance. Future directions entail the improvement of the labeled dataset, especially with regards to domain-specificity and incorporating context information. Additionally, other modalities and Natural Language Processsing (NLP)-based alliance assessment could be integrated.
Authors
- Jana Schaffrath (ORCID: https://orcid.org/0000-0003-0106-0243)
- Christopher Lalk (ORCID: https://orcid.org/0000-0002-1793-3507)
- Antonia Vehlen (ORCID: https://orcid.org/0000-0002-6019-3161)
- Steffen T. Eberhardt (ORCID: https://orcid.org/0000-0002-9900-4671)
- Julian Rubel (ORCID: https://orcid.org/0000-0002-9625-6611)
- Brian Schwartz (ORCID: https://orcid.org/0000-0003-4695-4953)
- Wolfgang Lutz (ORCID: https://orcid.org/0000-0002-5141-3847)
- Tobias Steinbrenner
- Kim Targan
Institutions
- Osnabrück University (DE)
- Universität Trier (DE)
Publication Details
- Journal
- Frontiers in Psychiatry
- Published
- 2025-05-09
- DOI
- https://doi.org/10.3389/fpsyt.2025.1504306
- Citations
- 5
- Primary Topic
- Mental Health via Writing
- Type
- article
- Field-Weighted Citation Impact
- 12.19
Funders
- Deutsche Forschungsgemeinschaft