Predicting Depression via Social Media

Прогнозирование депрессии с помощью социальных сетей
Eric Horvitz, Munmun De Choudhury, Scott Counts, Michael Gamon, Scott Counts
2021-08-03

Twitter usersmajor depressive disorderpsychometric instrumentsocial media detectionstatistical classifier
Major depression constitutes a serious challenge in personal and public health. Tens of millions of people each year suffer from depression and only a fraction receives adequate treatment. We explore the potential to use social media to detect and diagnose major depressive disorder in individuals. We first employ crowdsourcing to compile a set of Twitter users who report being diagnosed with clinical depression, based on a standard psychometric instrument. Through their social media postings over a year preceding the onset of depression, we measure behavioral attributes relating to social engagement, emotion, language and linguistic styles, ego network, and mentions of antidepressant medications. We leverage these behavioral cues, to build a statistical classifier that provides estimates of the risk of depression, before the reported onset. We find that social media contains useful signals for characterizing the onset of depression in individuals, as measured through decrease in social activity, raised negative affect, highly clustered egonetworks, heightened relational and medicinal concerns, and greater expression of religious involvement. We believe our findings and methods may be useful in developing tools for identifying the onset of major depression, for use by healthcare agencies; or on behalf of individuals, enabling those suffering from depression to be more proactive about their mental health.
1
A statistical classifier using social engagement, emotion, language, network, and medication-related cues estimates depression risk before reported onset.
2
Depression risk is associated with decreased social activity, increased negative affect, and more highly clustered ego networks.
3
Individuals approaching depression show heightened concerns related to relationships and antidepressant medications, alongside greater expression of religious involvement.
4
The findings suggest social media could support earlier identification of major depression by healthcare agencies or individuals, although the abstract does not report quantitative classifier performance.
5
Twitter postings contain behavioral signals that can help characterize the onset of major depression before individuals report its occurrence.

Twitter users who report clinically diagnosed major depressive disorder and their social media postings during the year preceding depression onset

Behavioral and linguistic social-media signals associated with and predictive of major depression onset, including changes in social activity, affect, ego-network structure, relational and medication concerns, and religious expression

Publication Details
Publication Date
2021-08-03
Journal
Publisher
ISSN
Cited by
1602
Access Type
Author Information
Authors
Eric Horvitz
Munmun De Choudhury
Scott Counts
Michael Gamon
Scott Counts
Explore further
Open the scid.ai AI chat with a ready-made request: it will find papers on a similar topic and help build a literature review.
Find similar papers in the chat →
Make a presentation
100%