On the State of NLP Approaches to Modeling Depression in Social Media: A Post-COVID-19 Outlook

TL;DR

This review surveys NLP approaches for modeling depression in social media with a post-COVID-19 perspective, synthesizing 231 papers published from 2020 to 2023 and identifying 38 COVID-19–focused studies. It tracks the field’s shift from feature-based classicalML to transformer-based architectures, including domain-specific models like MentalBERT and DisorBERT, and examines the emergence and limitations of LLMs in this domain. The paper discusses methodological issues such as semantic shift, model calibration, and ethical considerations, and catalogues datasets, many with restricted access or reliance on self-disclosed diagnoses. The authors highlight practical implications for real-time surveillance, data availability, and multilingual/multimodal research, and propose directions involving instruction-tuned LLMs, multimodal data, and broader language coverage to advance robust, responsible depression detection in social media.

Abstract

Computational approaches to predicting mental health conditions in social media have been substantially explored in the past years. Multiple reviews have been published on this topic, providing the community with comprehensive accounts of the research in this area. Among all mental health conditions, depression is the most widely studied due to its worldwide prevalence. The COVID-19 global pandemic, starting in early 2020, has had a great impact on mental health worldwide. Harsh measures employed by governments to slow the spread of the virus (e.g., lockdowns) and the subsequent economic downturn experienced in many countries have significantly impacted people's lives and mental health. Studies have shown a substantial increase of above 50% in the rate of depression in the population. In this context, we present a review on natural language processing (NLP) approaches to modeling depression in social media, providing the reader with a post-COVID-19 outlook. This review contributes to the understanding of the impacts of the pandemic on modeling depression in social media. We outline how state-of-the-art approaches and new datasets have been used in the context of the COVID-19 pandemic. Finally, we also discuss ethical issues in collecting and processing mental health data, considering fairness, accountability, and ethics.

Figures (2)

• Figure 1: Number of papers on depression modeling published during the pandemic (2020-2023) in peer-reviewed conferences or journals.
• Figure 2: Taxonomy of papers published between 2020-2023. Only 38 out of the 231 papers addressed depression modeling in the context of the pandemic.