Text-Based Detection and Understanding of Changes in Mental Health

Yaoyiran Li, Rada Mihalcea, Steven R. Wilson

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract / Description of output

Previous work has investigated the identification of mental health issues in social media users, yet the way that users' mental states and related behavior change over time remains relatively understudied. This paper focuses on online mental health communities and studies how users' contributions to these communities change over one year. We define a metric called the Mental Health Contribution Index (MHCI), which we use to measure the degree to which users' contributions to mental health topics change over a one-year period. In this work, we study the relationship between MHCI scores and the online expression of mental health symptoms by extracting relevant linguistic features from user-generated content and conducting statistical analyses. Additionally, we build a classifier to predict whether or not a user's contributions to mental health subreddits will increase or decrease. Finally, we employ propensity score matching to identify factors that correlate with an increase or a decrease in mental health forum contributions. Our work provides some of the first insights into detecting and understanding social media users' changes in mental health states over time.
Original languageEnglish
Title of host publicationInternational Conference on Social Informatics
EditorsSteffen Staab, Olessia Koltsova, Dmitry I. Ignatov
Place of PublicationCham
PublisherSpringer International Publishing AG
Number of pages13
ISBN (Electronic)978-3-030-01159-8
ISBN (Print)978-3-030-01158-1
Publication statusPublished - 20 Sept 2018
Event10th International Conference on Social Informatics 2018 - Saint Petersburg, Russian Federation
Duration: 25 Sept 201828 Sept 2018

Publication series

NameLecture Notes in Computer Science (LCNS)
PublisherSpringer, Cham
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Conference10th International Conference on Social Informatics 2018
Abbreviated titleSocInfo 2018
Country/TerritoryRussian Federation
CitySaint Petersburg
Internet address

Keywords / Materials (for Non-textual outputs)

  • Natural language processing
  • Mental health
  • Social media


Dive into the research topics of 'Text-Based Detection and Understanding of Changes in Mental Health'. Together they form a unique fingerprint.

Cite this