Psycholinguistic Analysis of Online Chat in Detecting Signs of Depression and Other Mental Health Issues

• Main Authors: Ross Azura, Zahit, Amalia, Madihie, Salmah, Mohamad Yusoff, Ida Juliana, Hutasuhut, Mohamad Azhari, Abu Bakar
• Format: Proceeding
• Language: English
• Published: 2022
• Subjects: H Social Sciences (General)
• Online Access: http://ir.unimas.my/id/eprint/40738/1/Psycholinguistic%20Analysis%20of%20Online%20Chat%20in%20Detecting%20Signs%20of%20Depression%20and%20Other%20Mental%20Health%20Issues.pdf; http://ir.unimas.my/id/eprint/40738/

In the last decades, people with mental health issues were referred to mental health professionals for diagnosis and treatment, forming stigmatisation among community members. Due to that stigma, new generations opt to seek help via online platforms. This study examines the online chat users' affective and psycholinguistic patterns, which could suggest signs of depression and other mental health issues. A randomised sample of 4000 chat items was extracted from the primary dataset generated from one of the online communities. This study was completed in two phases. In the first phase, categorisation was performed by five mental health-related professionals individually. The chat items were categorised based on the DSM-5 criteria of Major Depressive Disorder (MDD) and Beck Depression Inventory (BDI). In the second phase, inter-rater reliability and descriptive analysis were conducted. The result of inter-rater reliability for the depression classification ranged from good to strong value, with an average ρ=0.72. The chat items indicate one or more symptoms for the diagnosis of MDD, and other mental health issues (e.g. anxiety). Further evaluation should be conducted to understand the underlying meaning of each chat item so that holistic mental health care services and support could be offered to facilitate online communities.