Predicting Depression via Social Media

By M. de Choudhury, Michael Gamon, Scott Counts, Eric Horvitz. 2013

1,877 citations recorded by the source corpus.

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.

Subfields: Applied Psychology, Experimental and Cognitive Psychology, Social Psychology, Natural Language Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Applied Psychology

In Experimental and Cognitive Psychology

In Social Psychology

In Natural Language Processing

Citations in the corpus

Anticipating Safety Issues in E2E Conversational AI: Framework and Tooling (2021, background): “Intentions of self harm Amongst the large body of literature on depression detection and mental health assessment in social media (e.g., Benton et al., 2017; Coppersmith et al., 2014; De Choudhury et al., 2013, inter alia), some research focuses on detecting risk of self-harm.”
SafetyKit: First Aid for Measuring Safety in Open-domain Conversational Systems (2022, background): “Intentions of self harm Amongst the large body of work on mental health assessment in social media (e.g., Benton et al., 2017; Coppersmith et al., 2014; De Choudhury et al., 2013, inter alia), some research focuses on detecting risk of self-harm.”
Detecting Changes in Suicide Content Manifested in Social Media Following Celebrity Suicides (2015, background): “Suicidal ideation is associated with greater self-attentional focus [57, 60, 12].”
The Hitchhiker’s Guide to Computational Linguistics in Suicide Prevention (2021, background): “…depressive symptoms (e.g., sad, worthless, cry, lonely) as well as by more implicit language signals such as excessive usage of first-person pronouns (e.g., I ), past tense verbs (e.g., had ), somatic complaints, and hostility (De Choudhury et al., 2013; Eichstaedt et al., 2018; Rude et al., 2004).”
Emotional States vs. Emotional Words in Social Media (2015, neutral): “Human Factors, Measurement, Theory”
Leveraging Large Language Models for Mental Health Prediction via Online Text Data (2023, methodology): “For example, De Choudhury et al. [26] extracted a number of linguistic styles and other features to build an SVM model to perform depression prediction.”
Predictive Analysis on Twitter: Techniques and Applications (2018, methodology): “ Nw, [130] F1=0.59 CNN Stat, Txt [131] F1=0.83 User Classication RF Vis [49] F1=0.77 NB [132] Acc=0.82 LogR, LASSO [89] Acc=0.92 [90] F1=˘0.75 [114] F1=0.76 [54] AUC=˘0.7 [15] AUC=0.8 [108] Acc=0.84 [53] Acc=0.72 User Attitude, Personality, SVM, NB, RF TwM, Txt, Ling, [133] F1=0.62 and Mood Predicti”
The power of prediction with social media (2013, background): “In other words, one could exploit social media data as a kind of predictive market .”
Recognizing Depression from Twitter Activity (2015, methodology): “This approach was used to investigate the extent to which depression risk can be estimated for a population different from the population considered by the prior research [14].”
Quantifying Mental Health Signals in Twitter (2014, methodology): “Introspection here shows us what quantified signals in the content the classifiers base their decision on, and thus we can gain intuition about what signals are present in the content relevant to mental health.”
Natural language processing in mental health applications using non-clinical texts† (2017, methodology): “In another study, the same authors focused on measuring behavioral and emotional changes that could be indicative of postpartum depression (De Choudhury et al. 2013a).”
Suicide Ideation Detection via Social and Temporal User Representations using Hyperbolic Learning (2021, background): “Early efforts in leveraging NLP for suicide ideation detection on social media (De Choudhury et al., 2013, 2016; Shing et al., 2018; Sawhney et al., 2018) combine general features such as n-grams and POS tags with lexicons like LIWC (Pennebaker et al., 2001).”
A Social Media Study on the Effects of Psychiatric Medication Use (2019, background): “…singular and auxiliary verb shows negative coefficients, indicating that this drug might not be effective in those with greater preoccupation and self-attentional focus—the known characteristics of these two attribute usage, typically prevalent in depressed individuals (De Choudhury et al. 2013).”