A Psychologically Informed Part-of-Speech Analysis of Depression in Social Media

Proceedings of Recent Advances in Natural Language Processing, pages 199–207 Sep 1–3, 2021. https://doi.org/10.26615/978-954-452-072-4_024 199 A Psychologically Informed Part-of-Speech Analysis of Depression in Social Media Ana-Maria Bucur1 Ioana R. Podin˘a2 1 Interdisciplinary School of Doctoral Studies 2 Laboratory of Cognitive Clinical Sciences, Department of Psychology 3 Faculty of Mathematics and Computer Science University of Bucharest, Romania ana-maria.bucur@drd.unibuc.ro, ioana.podina@fpse.unibuc.ro ldinu@fmi.unibuc.ro Liviu P. Dinu3 Abstract In this work, we provide an extensive part-of- speech analysis of the discourse of social me- dia users with depression. Research in psy- chology revealed that depressed users tend to be self-focused, more preoccupied with them- selves and ruminate more about their lives and emotions. Our work aims to make use of large- scale datasets and computational methods for a quantitative exploration of discourse. We use the publicly available depression dataset from the Early Risk Prediction on the Inter- net Workshop (eRisk) 2018 and extract part- of-speech features and several indices based on them. Our results reveal statistically sig- nificant differences between the depressed and non-depressed individuals confirming findings from the existing psychology literature. Our work provides insights regarding the way in which depressed individuals are expressing themselves on social media platforms, allow- ing for better-informed computational models to help monitor and prevent mental illnesses. 1 Introduction Mental health disorders are a common problem in our world. Currently, mental health issues are on the rise: there is a 13% increase in the past decade according to World Health Organization (WHO)1, with depression being at the forefront. Many mental illnesses remain undiagnosed due to social stigma, leading people to live 1 in 5 years of disability in their lifetime. With the rise of social media websites, interdisci- plinary researchers in natural language processing, psychology and network analysis have turned their attention to automatically detect and monitor men- tal health manifestations through users’ individual activity on social platforms (e.g. Facebook, Twit- ter, Reddit). The research is primarily focused on 1https://www.who.int analyzing users’ texts from posts and comments and determining, through computational linguis- tics models, the risk for various mental conditions - self-harm, depression, addictions etc. Research is fulled through curated datasets (Yates et al., 2017; Losada and Crestani, 2016; Amir et al., 2019) with texts from primarily Reddit and Twitter. At the forefront of incentivising inter- disciplinary research on monitoring mental health on social media are workshops such as the Early risk prediction on the Internet (eRisk) Workshop2 and the Workshop on Computational Linguistics and Clinical Psychology (CLPsych)3. CLPsych and eRisk are two significant initia- tives focusing on the interdisciplinary research area between computational linguistics and psychology. The eRisk Workshop, from the Conference and Labs of the Evaluation Forum (CLEF), focuses on the technologies that can be used for early risk de- tection of different pathologies or safety threats (Losada et al.). In five years of existence, the work- shop addressed multiple mental health problems: pathological gambling, depression, self-harm and anorexia. The CLPsych Workshop was co-located with several international conferences on natural lan- guages processing, the last edition was …

Content truncated to 500 words