Leveraging Mental Health Forums for User-level Depression Detection on Social Media

Proceedings of the 13th Conference on Language Resources and Evaluation (LREC 2022), pages 5418–5427 Marseille, 20-25 June 2022 © European Language Resources Association (ELRA), licensed under CC-BY-NC-4.0 5418 Leveraging Mental Health Forums for User-level Depression Detection on Social Media Sravani Boinepelli1, Tathagata Raha1, Harika Abburi1, Pulkit Parikh1, Niyati Chhaya2, Vasudeva Varma1 1IIIT-Hyderabad Information Retrieval and Extraction Lab, Gachibowli, Hyderabad, Telangana, India sravani.boinepelli,tathagata.raha,harika.a,pulkit.parikh@research.iiit.ac.in, vv@iiit.ac.in 2Adobe Research Bengaluru, Karnataka, India nchhaya@adobe.com Abstract The number of depression and suicide risk cases on social media platforms is ever-increasing, and the lack of depression detection mechanisms on these platforms is becoming increasingly apparent. A majority of work in this area has focused on leveraging linguistic features while dealing with small-scale datasets. However, one faces many obstacles when factoring into account the vastness and inherent imbalance of social media content. In this paper, we aim to optimize the performance of user-level depression classification to lessen the burden on computational resources. The resulting system executes in a quicker, more efficient manner, in turn making it suitable for deployment. To simulate a platform agnostic framework, we simultaneously replicate the size and composition of social media to identify victims of depression. We systematically design a solution that categorizes post embeddings, obtained by fine-tuning transformer models such as RoBERTa, and derives user-level representations using hierarchical attention networks. We also introduce a novel mental health dataset to enhance the performance of depression categorization. We leverage accounts of depression taken from this dataset to infuse domain-specific elements into our framework. Our proposed methods outperform numerous baselines across standard metrics for the task of depression detection in text. Keywords: Depression Detection, Text classification, Semi-supervised, weakly-supervised and unsupervised learning, Social Media Processing 1. Introduction Depression is viewed as the largest contributor to global disability and often co-occurs with anxiety or other psychological and physical disorders. One of the many risks of clinical depression is suicide. Re- search has indicated that approximately two-thirds of people who die by suicide were dealing with depres- sion at the time of death (Richards and OHara, 2014). Meanwhile, according to the World Health Organiza- tion, more than 300 million people from varied de- mographics suffer from depression and nearly 50% of people worldwide with clinical depression remain un- treated (WHO, 2017). Over the last couple of years, there has been an influx of depression and mental health posts on social me- dia which has only been exacerbated by the COVID- 19 pandemic. The abundance of abusive language on the web has also contributed to the profusion of men- tal health and depression-related issues, especially in teens, making life online increasingly hazardous (Fer- sko, 2018; Karim et al., 2020). In this paper, we take steps towards identifying not just singular accounts of depression (post-level classification), but towards de- tection of individuals suffering from depression (user- level classification). By post-level classification, we refer to the labelling of isolated social media posts whereas user-level classification labels a user based on a collection of posts or post history of the user. Our goal …

Content truncated to 500 words