@inproceedings{alqahtani-etal-2026-stressroberta,
    title = "{S}tress{R}o{BERT}a: Cross-Condition Transfer Learning from Depression, Anxiety, and {PTSD} to Stress Detection",
    author = "Alqahtani, Amal Abdullah  and
      Kayi, Efsun  and
      Diab, Mona T.",
    editor = {Danilova, Vera  and
      Kurfal{\i}, Murathan  and
      S{\"o}derfeldt, Ylva  and
      Reed, Julia  and
      Burchell, Andrew},
    booktitle = "Proceedings of the 1st Workshop on Linguistic Analysis for Health ({H}ea{L}ing 2026)",
    month = mar,
    year = "2026",
    address = "Rabat, Morocco",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-eacl/2026.healing-1.27/",
    pages = "305--313",
    ISBN = "979-8-89176-367-8",
    abstract = "The prevalence of chronic stress represents a major public health concern, yet automated detection of vulnerable individuals remains limited. Social media platforms like X (formerly Twitter) serve as important venues for people to share their experiences openly. This paper introduces StressRoBERTa, a cross-condition transfer learning approach for the automatic detection of self-reported chronic stress in English tweets. We investigate whether continual pretraining on clinically related conditions, such as depression, anxiety, and PTSD, which have a high comorbidity with chronic stress, improves stress detection compared to general language models. We continually pretrained RoBERTa on the Stress-SMHD corpus, a subset of Self-reported Mental Health Diagnoses focused on stress-related conditions, consisting of 108 million words from users with self-reported diagnoses of depression, anxiety, and PTSD. Then, we fine-tuned on the SMM4H 2022 Shared Task 8. StressRoBERTa achieves 82{\%} F1, which outperforms the best shared task system (79{\%} F1) by 3 percentage points. Our results demonstrate that focused cross-condition transfer learning from stress-related disorders provides stronger representations than general mental health training. To validate cross-condition generalization, we also fine-tuned the model on the Dreaddit. Our result of 81{\%} F1 further demonstrates the transfer from clinical mental health contexts to situational stress discussions."
}