{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/quantifying-mental-health-from-social-media","title":"Quantifying Mental Health from Social Media with Neural User Embeddings","arxiv_id":"1705.00335","date":"2017-04-30","proceeding":null,"authors":["Silvio Amir","Glen Coppersmith","Paula Carvalho","Mário J. Silva","Byron C. Wallace"],"abstract":"Mental illnesses adversely affect a significant proportion of the population\nworldwide. However, the methods traditionally used for estimating and\ncharacterizing the prevalence of mental health conditions are time-consuming\nand expensive. Consequently, best-available estimates concerning the prevalence\nof mental health conditions are often years out of date. Automated approaches\nto supplement these survey methods with broad, aggregated information derived\nfrom social media content provides a potential means for near real-time\nestimates at scale. These may, in turn, provide grist for supporting,\nevaluating and iteratively improving upon public health programs and\ninterventions.\n  We propose a novel model for automated mental health status quantification\nthat incorporates user embeddings. This builds upon recent work exploring\nrepresentation learning methods that induce embeddings by leveraging social\nmedia post histories. Such embeddings capture latent characteristics of\nindividuals (e.g., political leanings) and encode a soft notion of homophily.\nIn this paper, we investigate whether user embeddings learned from twitter post\nhistories encode information that correlates with mental health statuses. To\nthis end, we estimated user embeddings for a set of users known to be affected\nby depression and post-traumatic stress disorder (PTSD), and for a set of\ndemographically matched `control' users. We then evaluated these embeddings\nwith respect to: (i) their ability to capture homophilic relations with respect\nto mental health status; and (ii) the performance of downstream mental health\nprediction models based on these features. Our experimental results demonstrate\nthat the user embeddings capture similarities between users with respect to\nmental conditions, and are predictive of mental health.","url_abs":"http://arxiv.org/abs/1705.00335v1","url_pdf":"http://arxiv.org/pdf/1705.00335v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"quantifying-mental-health-from-social-media","repo_url":"https://github.com/samiroid/usr2vec","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"representation-learning","task_name":"Representation Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}