Elamur Research⌕
☰
home / papers / 2026 / Understanding the Sociocultural Dimensions of Mental Health Discourse in Arabic X Communities

Understanding the Sociocultural Dimensions of Mental Health Discourse in Arabic X Communities

Paper · 2026 · Amal Abdullah Alqahtani, Rana Aref Salama and Mona T. Diab

Summary

Most computational work on mental health in social media studies English speakers. This paper turns to Arabic and describes how people talk about three conditions in condition-specific X Communities: BPD, Bipolar and ADHD. The authors avoid diagnosing anyone. They frame the work as descriptive computational social science.

They collected about ten thousand posts and cleaned them. GPT-4.1 then filtered the posts to authors who appear to write from lived experience, with Qwen3 run in parallel as a cross-check and a small human-annotated sample for validation. The filtered corpus was analysed with:

  • time-of-day activity profiles
  • weighted log-odds
  • NMF topics
  • code-switching counts
  • a six-domain cultural keyword scheme based on Kleinman's explanatory models of illness

Each community leans toward different vocabulary:

  • Bipolar: religious and medical terms
  • BPD: relational, identity and distress language
  • ADHD: practical talk about symptoms and medication

The authors present these contrasts only as hypotheses. The corpus is unbalanced and partly concentrated in time, and the keyword scheme has not been validated. The main reusable outputs are the disclosure-filtering pipeline and the keyword framework.

Key points

  • Corpus size: 10,091 raw tweets were reduced to 9,582 after preprocessing (1,286 users, 98.4% Arabic). Disclosure filtering then left 8,147 tweets from 607 users, collected between March 2022 and February 2026.
  • Annotation setup: GPT-4.1 was the primary annotator and Qwen3-235B-A22B the screening model, both at temperature 0. The prompt included the user bio, a bio override rule, a default of Negative when unsure, and a 13-tag reason taxonomy.
  • Reliability:
    • The two models agreed with each other at κ = 0.84 (90.8% consensus).
    • Agreement with humans was much lower: κ = 0.631 for GPT-4.1 and κ = 0.329 for Qwen3.
    • The human set had 200 tweets, and the two annotators agreed on 192 of them.
  • Code-switching: content-bearing English appears in 28.5% of ADHD tweets, against 6.3% for BPD and 3.2% for Bipolar. Of the ADHD tweets, 54 contain the acronym "ADHD" itself.
  • Imbalance and co-occurrence:
    • The ADHD subcorpus has only 253 tweets.
    • 83.5% of BPD tweets come from one nine-month window.
    • 10.3% of Bipolar tweets contain both religious and medical keywords.
    • Only X Post IDs are released.

Interesting ideas

  • I1. Two LLM annotators run in parallel, and their disagreements mark ambiguous items. The paper also shows that agreement between the models greatly overstates reliability compared with agreement against humans, so a small human gold set stays necessary.
  • I2. The annotation prompt makes the model attach a reason tag from a fixed taxonomy, and it falls back to a conservative default label when evidence is weak. This makes LLM labels auditable and keeps their errors leaning one way.
  • I3. Author metadata (the bio) is used as context, with an explicit override rule for self-identification that a single post leaves unclear.
  • I4. Pluralism is measured within single texts, by counting tweets that combine two domains, instead of comparing aggregate rates per domain. Within-text counts are better evidence about what an individual author thinks.
  • I5. Code-switching gets a simple, reproducible operational definition: remove placeholder tokens, then count Latin-script tokens of at least two characters that are not stopwords. Small subsets get a binomial confidence interval.

For Mova

How these ideas could serve Mova, our small open language model — its files, stages and tests:

  • I1 (learn): when several witness parsers agree on silver data, Mova should still compare that agreement against a hand-checked sample, rather than treating consensus alone as the measure of reliability.
  • I2 (learn): LLM teacher tiers could return a tag from a closed list explaining each correction and abstain by default, so that repair rules can be grouped and audited by reason.
  • I3 (frames): speaker identity statements elsewhere in a document could feed speech-act and stance interpretation, so that first-person claims are attributed correctly.
  • I4 (world): when inferring an author's beliefs, Mova should look at which concepts co-occur inside one text's world model, not at corpus-level frequencies.
  • I5 (tokens): a script-based flag for foreign tokens, with placeholders removed first, could route mixed-language spans to a separate tagger path, and rates on small subsets should be reported with intervals.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🎓Carnegie Mellon UniversityPittsburgh · United States539 works · 179,533 citations🎓Cairo UniversityGiza · Egypt42 works · 1,350 citations🎓King Saud UniversityRiyadh · Saudi Arabia37 works · 527 citations

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksPsychology · 1,552 worksSocial Psychology · 306 works

🧭 Topics 2Research subjects and languages the work is about.Semitic languages · 2,908 worksDiscourse analysis · 466 works

Cite

Cite as Amal Abdullah Alqahtani, Rana Aref Salama and Mona T. Diab. 2026. Understanding the Sociocultural Dimensions of Mental Health Discourse in Arabic X Communities. Proceedings of the 11th Social Media Mining for Health Research and Applications (SMM4H-HeaRD 2026) Workshop and Shared Tasks.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.smm4h-1.47

@inproceedings{alqahtani2026understanding,
  title = {{Understanding the Sociocultural Dimensions of Mental Health Discourse in Arabic X Communities}},
  author = {Amal Abdullah Alqahtani, Rana Aref Salama and Mona T. Diab},
  booktitle = {Proceedings of the 11th Social Media Mining for Health Research and Applications (SMM4H-HeaRD 2026) Workshop and Shared Tasks},
  year = {2026},
  doi = {10.18653/v1/2026.smm4h-1.47},
  url = {https://doi.org/10.18653/v1/2026.smm4h-1.47}
}

References

26 references, 6 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Muhammad Abdul-Mageed, AbdelRahim Elmadany, and El Moatez Billah Nagoudi. 2021. ARBERT & MARBERT: Deep bidirectional transformers for Arabic. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers), pages 7088–7105, Online. Association for Computational Linguistics. ↗
  2. Horeya AbouWarda, Mateusz Dolata, and Gerhard Schwabe. 2024. How does an online mental health community on Twitter empower diverse population levels and groups? a qualitative analysis of #Bipo-larClub. Journal of Medical Internet Research, 26:e55965. find
  3. Norah Mohammed Alamri. 2022. Arabic-English code-switching among KKU students on social media Twitter. M.A. thesis, English Department, Faculty of Languages and Translation, King Khalid University; hosted by Arab World English Journal, ID No. 284. ↗
  4. Jamal M Alkhateeb and Muna S Alhadidi. 2019. ADHD research in Arab countries: A systematic review of literature. Journal of attention disorders, 23(13):1531–1545. find
  5. Mohammed M. J. Alqahtani, Nouf Mohammed Al Saud, Nawal Mohammed Alsharef, Ahmad N. AlHadi, Saleh Mohammed Alsalhi, Elham H. Al-Hifthy, Yasser Ad-Dab’bagh, Nader Alrahili, Fawwaz Abdul-razaq Alenazi, Barakat M. Alotaibi, Sultan Mahmoud Alsaeed, Boshra A. Arnout, Latifah ALQasem, Abdulkarim Alhossein, Yasser Jubran Alqahtani, Sami-rah A. AlGhamdi, Jeremy Varnham, Saeed Abdulwahab Asiri, and Maysaa W. Buraik. 2025. Standardization of the Arabic version of the adult ADHD self-report screening scale for DSM-5 (ASRS-5) among adults in Saudi Arabia: Variability of ADHD screening according to sociodemographic variables. Journal of Attention Disorders, 29(6):445–457. ↗
  6. Wissam Antoun, Fady Baly, and Hazem Hajj. 2020. AraBERT: Transformer-based model for Arabic language understanding. In Proceedings of the 4th Workshop on Open-Source Arabic Corpora and Processing Tools, with a Shared Task on Offensive Language Detection, pages 9–15, Marseille, France. European Language Resources Association. ↗
  7. Lama Ayash, Ashwag Alasmari, and Hassan Alhuzali. 2025. ContextMentalQA: Modeling cultural, social, and religious context in Arabic mental health question answering. TechRxiv Preprint. ↗
  8. Glen Coppersmith, Mark Dredze, and Craig Harman. 2014. Quantifying mental health signals in Twitter. In Proceedings of the workshop on computational linguistics and clinical psychology: From linguistic signal to clinical reality, pages 51–60.
  9. Glen Coppersmith, Mark Dredze, Craig Harman, and Kristy Hollingshead. 2015. From ADHD to SAD: Analyzing the language of mental health on Twitter through self-reported diagnoses. In Proceedings of the 2nd workshop on computational linguistics and clinical psychology: from linguistic signal to clinical reality, pages 1–10. find
  10. Erving Goffman. 1974. Frame Analysis: An Essay on the Organization of Experience. Harper & Row, New York. find
  11. Arthur Kleinman. 1980. Patients and healers in the context of culture: An exploration of the borderland between anthropology, medicine, and psychiatry, volume 3. Univ of California Press. find
  12. J. Richard Landis and Gary G. Koch. 1977. The measurement of observer agreement for categorical data. Biometrics, 33(1):159–174. ↗
  13. David Lazer, Alex Pentland, Lada Adamic, Sinan Aral, Albert-Laszlo Barabasi, Devon Brewer, Nicholas Christakis, Noshir Contractor, James Fowler, Myron Gutmann, Tony Jebara, Gary King, Michael Macy, Deb Roy, and Marshall Van Alstyne. 2009. Computational social science. Science, 323(5915):721–723. ↗
  14. Daniel D Lee and H Sebastian Seung. 1999. Learning the parts of objects by non-negative matrix factorization. Nature, 401(6755):788–791. ↗
  15. Burt L. Monroe, Michael P. Colaresi, and Kevin M. Quinn. 2008. Fightin’ words: Lexical feature selection and evaluation for identifying the content of political conflict. Political Analysis, 16(4):372–403. ↗
  16. Latefa A. Dardas and Leigh Ann Simmons. 2015. The stigma of mental illness in Arab families: A concept analysis. Journal of Psychiatric and Mental Health Nursing, 22(9):668–679. find
  17. Carol Myers-Scotton. 1997. Duelling Languages: Grammatical Structure in Codeswitching. Oxford University Press, Oxford. Reprint/edition; originally published in 1993. find
  18. Munmun De Choudhury, Michael Gamon, Scott Counts, and Eric Horvitz. 2013. Predicting depression via social media. In Proceedings of the international AAAI conference on web and social media, volume 7, pages 128–137. find
  19. Bosheng Ding, Chengwei Qin, Linlin Liu, Yew Ken Chia, Boyang Li, Shafiq Joty, and Lidong Bing. 2023. Is GPT-3 a good data annotator? In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 11173–11195, Toronto, Canada. Association for Computational Linguistics. ↗
  20. Mario Eid, Venise Abi Kheir, Maya Bizri, Amine Lar-naout, and Samer El Hayek. 2025. Somatic symptom and related disorders in the Arab world: a narrative review of clinical features and care implications. Frontiers in Psychiatry, 16:1692267. ↗
  21. Sindhu Kiranmai Ernala, Michael L Birnbaum, Kristin A Candan, Asra F Rizvi, William A Sterling, John M Kane, and Munmun De Choudhury. 2019. Methodological gaps in predicting mental health states from social media: Triangulating diagnostic signals. In Proceedings of the 2019 chi conference on human factors in computing systems, pages 1–16. find
  22. Fabrizio Gilardi, Meysam Alizadeh, and Maël Kubli. 2023. ChatGPT outperforms crowd workers for text-annotation tasks. Proceedings of the National Academy of Sciences, 120(30):e2305016120. ↗
  23. Erving Goffman. 1959. The Presentation of Self in Everyday Life. Doubleday, Garden City, New York. find
  24. Kailai Yang, Shaoxiong Ji, Tianlin Zhang, Qianqian Xie, Ziyan Kuang, and Sophia Ananiadou. 2023. Towards interpretable mental health analysis with large language models. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 6056–6077, Singapore. Association for Computational Linguistics. find
  25. Wajdi Zaghouani, Eman Sedqy Shlkamy, and Mabrouka Bessghaier. 2026. From posts to pressure: An Arabic dataset about stress and mental-health monitoring. In Proceedings of the 2nd Workshop on NLP for Languages Using Arabic Script, pages 422–432, Rabat, Morocco. Association for Computational Linguistics.
  26. Monica Zolezzi, Maha Alamri, Shahd Shaar, and Daniel Rainkie. 2018. Stigma associated with mental illness and its treatment in the Arab culture: A systematic review. International Journal of Social Psychiatry, 64(6):597–609. ↗

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/alqahtani-2026-understanding-sociocultural-dimensions-mental-health/