Semi-Supervised Approach to Monitoring Clinical Depressive Symptoms in Social Media

With the rise of social media, millions of people are routinely expressing their moods, feelings, and daily struggles with mental health issues on social media platforms like Twitter. Unlike traditional observational cohort studies conducted through questionnaires and self-reported surveys, we explore the reliable detection of clinical depression from tweets obtained unobtrusively. Based on the analysis of tweets crawled from users with self-reported depressive symptoms in their Twitter profiles, we demonstrate the potential for detecting clinical depression symptoms which emulate the PHQ-9 questionnaire clinicians use today. Our study uses a semi-supervised statistical model to evaluate how the duration of these symptoms and their expression on Twitter (in terms of word usage patterns and topical preferences) align with the medical findings reported via the PHQ-9. Our proactive and automatic screening tool is able to identify clinical depressive symptoms with an accuracy of 68% and precision of 72%.

[1]  Naomie Salim,et al.  Fuzzy Based Implicit Sentiment Analysis on Quantitative Sentences , 2017, ArXiv.

[2]  R. Spitzer,et al.  The PHQ-9 , 2001, Journal of General Internal Medicine.

[3]  Eric Horvitz,et al.  Social media as a measurement tool of depression in populations , 2013, WebSci.

[4]  Amit P. Sheth,et al.  Analyzing Clinical Depressive Symptoms in Twitter , 2016 .

[5]  Li Sun,et al.  A Depression Detection Model Based on Sentiment Analysis in Micro-blog Social Network , 2013, PAKDD Workshops.

[6]  Zhengxing Huang,et al.  On mining latent topics from healthcare chat logs , 2016, J. Biomed. Informatics.

[7]  Yair Neuman,et al.  Proactive screening for depression through metaphorical and automatic text analysis , 2012, Artif. Intell. Medicine.

[8]  Mark Stevenson,et al.  Evaluating Topic Coherence Using Distributional Semantics , 2013, IWCS.

[9]  Leonardo Max Batista Claudino,et al.  Beyond LDA: Exploring Supervised Topic Modeling for Depression-Related Language in Twitter , 2015, CLPsych@HLT-NAACL.

[10]  Arjun Mukherjee,et al.  Aspect Extraction through Semi-Supervised Modeling , 2012, ACL.

[11]  M. Haselton,et al.  The Evolution of Cognitive Bias , 2015 .

[12]  Mark Dredze,et al.  Shared Task : Depression and PTSD on Twitter , 2015 .

[13]  Dirk Hovy,et al.  Multitask Learning for Mental Health Conditions with Limited Social Media Data , 2017, EACL.

[14]  Naomie Salim,et al.  Recognition of side effects as implicit-opinion words in drug reviews , 2016, Online Inf. Rev..

[15]  Timothy Baldwin,et al.  Automatic Evaluation of Topic Coherence , 2010, NAACL.

[16]  Paul Thompson,et al.  Predicting military and veteran suicide risk: Cultural aspects , 2014, CLPsych@ACL.

[17]  Philip S. Yu,et al.  Mining Online Social Data for Detecting Social Network Mental Disorders , 2016, WWW.

[18]  Ramesh Nallapati,et al.  Labeled LDA: A supervised topic model for credit attribution in multi-labeled corpora , 2009, EMNLP.

[19]  Rafael A. Calvo,et al.  CLPsych 2016 Shared Task: Triaging content in online peer-support forums , 2016, CLPsych@HLT-NAACL.

[20]  Xiaojin Zhu,et al.  Incorporating domain knowledge into topic modeling via Dirichlet Forest priors , 2009, ICML '09.

[21]  Eric Horvitz,et al.  Predicting Depression via Social Media , 2013, ICWSM.

[22]  David W. McDonald,et al.  Perception Differences between the Depressed and Non-Depressed Users in Twitter , 2013, ICWSM.

[23]  Xiaojin Zhu,et al.  Latent Dirichlet Allocation with Topic-in-Set Knowledge , 2009, HLT-NAACL 2009.

[24]  Susan T. Dumais,et al.  Partially labeled topic models for interpretable text mining , 2011, KDD.

[25]  Xiaohui Yan,et al.  A biterm topic model for short texts , 2013, WWW.

[26]  Mark Dredze,et al.  From ADHD to SAD: Analyzing the Language of Mental Health on Twitter through Self-Reported Diagnoses , 2015, CLPsych@HLT-NAACL.

[27]  Mark Dredze,et al.  Quantifying Mental Health Signals in Twitter , 2014, CLPsych@ACL.

[28]  Geoff Holmes,et al.  Classifier chains for multi-label classification , 2009, Machine Learning.

[29]  Glen Coppersmith,et al.  Exploratory Analysis of Social Media Prior to a Suicide Attempt , 2016, CLPsych@HLT-NAACL.

[30]  Maarten Sap,et al.  The role of personality, age, and gender in tweeting about mental illness , 2015, CLPsych@HLT-NAACL.

[31]  Glen Coppersmith,et al.  Quantifying the Language of Schizophrenia in Social Media , 2015, CLPsych@HLT-NAACL.

[32]  Andrew McCallum,et al.  Optimizing Semantic Coherence in Topic Models , 2011, EMNLP.

[33]  Xiaojin Zhu,et al.  Proceedings of the Twenty-Second International Joint Conference on Artificial Intelligence A Framework for Incorporating General Domain Knowledge into Latent Dirichlet Allocation Using First-Order Logic , 2022 .

[34]  Amit P. Sheth,et al.  Challenges of Sentiment Analysis for Dynamic Events , 2017, IEEE Intelligent Systems.

[35]  Svetha Venkatesh,et al.  Affective and Content Analysis of Online Depression Communities , 2014, IEEE Transactions on Affective Computing.

[36]  Thomas Wetter,et al.  Screening Internet forum participants for depression symptoms by assembling and enhancing multiple NLP methods , 2015, Comput. Methods Programs Biomed..