Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2412.20741 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909444022992896 |
|---|---|
| author | Rutowski, Tomasz Shriberg, Elizabeth Harati, Amir Lu, Yang Chlebek, Piotr Oliveira, Ricardo |
| author_facet | Rutowski, Tomasz Shriberg, Elizabeth Harati, Amir Lu, Yang Chlebek, Piotr Oliveira, Ricardo |
| contents | Digital screening and monitoring applications can aid providers in the management of behavioral health conditions. We explore deep language models for detecting depression, anxiety, and their co-occurrence from conversational speech collected during 16k user interactions with an application. Labels come from PHQ-8 and GAD-7 results also collected by the application. We find that results for binary classification range from 0.86 to 0.79 AUC, depending on condition and co-occurrence. Best performance is achieved when a user has either both or neither condition, and we show that this result is not attributable to data skew. Finally, we find evidence suggesting that underlying word sequence cues may be more salient for depression than for anxiety. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2412_20741 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Depression and Anxiety Prediction Using Deep Language Models and Transfer Learning Rutowski, Tomasz Shriberg, Elizabeth Harati, Amir Lu, Yang Chlebek, Piotr Oliveira, Ricardo Computation and Language Digital screening and monitoring applications can aid providers in the management of behavioral health conditions. We explore deep language models for detecting depression, anxiety, and their co-occurrence from conversational speech collected during 16k user interactions with an application. Labels come from PHQ-8 and GAD-7 results also collected by the application. We find that results for binary classification range from 0.86 to 0.79 AUC, depending on condition and co-occurrence. Best performance is achieved when a user has either both or neither condition, and we show that this result is not attributable to data skew. Finally, we find evidence suggesting that underlying word sequence cues may be more salient for depression than for anxiety. |
| title | Depression and Anxiety Prediction Using Deep Language Models and Transfer Learning |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2412.20741 |