Studying the language of mental illness in Romanian social media

Abstract

Mental health represents an important topic in the discussion about public health, as the prevalence of disorders increases, and the consequences are harder to manage. One way to address this problem is the use of computational methods to process social media data that provides valuable insight into individuals’ everyday activity. Particularly, there has been a focus on studying the language of mental health in textual content, as word choice, discourse structure and style have significant psychological implications. This paper aims to quantify the language of mental disorders in a set of posts and comments collected from a Romanian medical forum and explore illness specific characteristics through classification tasks. Results suggest that both vocabulary use and form of textual expression are largely discriminative for individual mental disorders.