Mental health metrics, namely psychological stress, lonely expressions, anxiety, and sentiment are measured daily using pre-trained machine learning models applied to a random 1% Twitter data. For more details, read our publication in the Journal of General Internal Medicine: http://wwbp.org/papers/jgim-2020.pdf
Dataset is available (upon request) at: https://zenodo.org/record/4064043
Data snapshot:
| group_id | feat | value | group_norm | day | cnty |
|---|---|---|---|---|---|
| 2020-04-16:01001 | lonely_score | 78 | 2.92268402613488 | 2020-04-16 | 01001 |
| 2020-04-16:01003 | lonely_score | 830 | 2.82928758282208 | 2020-04-16 | 01003 |
| 2020-04-16:01005 | lonely_score | 13 | 3.4083486715075 | 2020-04-16 | 01005 |
| 2020-04-16:01017 | lonely_score | 93 | 2.67820445675611 | 2020-04-16 | 01017 |
| 2020-04-16:01021 | lonely_score | 96 | 3.02387743147066 | 2020-04-16 | 01021 |
cnty: FIPS code of county
day: date
group_norm: mental health estimate; a sum of term relative frequencies weighted by their association with this mental health outcome in the pre-trained model
value: number of words contributing to the estimate
feat: descriptor of metric
group_id: concatenation of day:cnty
This data (aggregated to the state-level) is also used to update the Penn COVID Twitter Map https://penncovid19hub.com/twitter-map
APA:
Guntuku, S. C., Sherman, G., Stokes, D. C., Agarwal, A. K., Seltzer, E., Merchant, R. M., & Ungar, L. H. (2020). Tracking Mental Health and Symptom Mentions on Twitter During COVID-19. Journal of general internal medicine, 1-3.
Bib:
@article{guntuku2020tracking,
title={Tracking Mental Health and Symptom Mentions on Twitter During COVID-19},
author={Guntuku, Sharath Chandra and Sherman, Garrick and Stokes, Daniel C and Agarwal, Anish K and Seltzer, Emily and Merchant, Raina M and Ungar, Lyle H},
journal={Journal of general internal medicine},
pages={1--3},
year={2020},
publisher={Springer}
}
Details of sentiment, stress, and loneliness models are described in the following papers: Sentiment:
Mohammad, S. M., & Turney, P. D. (2013). Crowdsourcing a word–emotion association lexicon. Computational Intelligence, 29(3), 436-465.
Stress:
Guntuku, S. C., Buffone, A., Jaidka, K., Eichstaedt, J. C., & Ungar, L. H. (2019, July). Understanding and measuring psychological stress using social media. In Proceedings of the International AAAI Conference on Web and Social Media (Vol. 13, No. 01, pp. 214-225).
Loneliness:
Guntuku, S. C., Schneider, R., Pelullo, A., Young, J., Wong, V., Ungar, L., ... & Merchant, R. (2019). Studying expressions of loneliness in individuals using twitter: an observational study. BMJ open, 9(11).
Anxiety:
Schwartz HA, Eichstaedt J, Kern M, Park G, Sap M, Stillwell D, Kosinski M, Ungar L. Towards assessing changes in degree of depression through Facebook. In Proceedings of the workshop on computational linguistics and clinical psychology: from linguistic signal to clinical reality. Baltimore: Association for Computational Linguistics; 2014. p. 118–125.
Guntuku, S. C., Preotiuc-Pietro, D., Eichstaedt, J. C., & Ungar, L. H. (2019, July). What twitter profile and posted images reveal about depression and anxiety. In Proceedings of the international AAAI conference on web and social media (Vol. 13, pp. 236-246).
For any queries, please reach out at sharathg at cis dot upenn dot edu or garricks at sas dot upenn dot edu.