Related Experiment Video
Updated: Oct 3, 2025

Integrating Computerized Linguistic and Social Network Analyses to Capture Addiction Recovery Capital in an Online Community
Published on: May 31, 2019
Using language in social media posts to study the network dynamics of depression longitudinally
Sean W Kelley1,2, Claire M Gillan3,4,5
1School of Psychology, Trinity College Dublin, Dublin, Ireland. sekelley@tcd.ie.
This study found that the connections between language used in social media posts can indicate depression severity. Higher network connectivity in language features correlated with more severe depressive symptoms and episodes.
Area of Science:
- Psychology
- Computational Linguistics
- Network Science
Background:
- Mental illness can be understood through network theory, where symptoms interact causally.
- Depression network connectivity is increasingly recognized as a risk factor for developing and maintaining depressive states.
Purpose of the Study:
- To investigate the relationship between linguistic network connectivity and depression severity using social media data.
- To explore dynamic changes in network connectivity during depressive episodes.
Main Methods:
- Analysis of Twitter data from 946 participants who self-reported depressive episode dates and severity.
- Construction of personalized, within-subject networks based on depression-related linguistic features.
- Examination of associations between network connectivity metrics and depression severity.
Main Results:
- An association was found between current depression severity and 8 out of 9 examined text features.
- Individuals with higher depression severity exhibited greater overall network connectivity between depression-relevant linguistic features.
- Within-subject changes in overall network connectivity were observed, correlating with self-reported depressive episodes.
Conclusions:
- Personalized linguistic networks show dynamic changes in connectivity that reflect current depression symptoms.
- Network connectivity of depression-associated linguistic features may serve as a potential biomarker for depression severity and fluctuations.
Related Concept Videos
Long-term Depression
Depressive Disorders: Etiology
Biological Factors in Depression
Biological predispositions significantly influence the risk of developing depressive disorders. Genetic studies highlight the role of variations in the serotonin transporter...
Depressive Disorders: MDD and Dysthymia
Depression: Overview
Longitudinal Studies
Social Foundations of Self IV: Self in Digital Communication

