Effective Techniques for Depression Detection on Social Media: A Comprehensive Review

Currently depression has become one of the common and major public health problems in our society due to changes in lifestyle, high work pressure, persistent illness, loss of beloved ones in life, social and economic crisis due to pandemics, etc. Doctors have traditionally emphasized face-to-face interaction to diagnose depression using clinical depression criteria. Not every patient (less than 30^%) would seek medical care in the early stages of depression, and moreover many of them would not accept that they have a mental disorder, leading to delays in remedial measures, and deterioration of symptoms. Failure to detect depression in the early stages has led to suicide in many cases. To address this problem, and take advantage of the abundance of data, machine learning has been used significantly for depression detection in social networks. A variety of machine learning models, features, and media have been used in this area. After data pre-processing, Feature extraction techniques are used to extract psychological factors, Linguistic factors, and emotional factors from Text (LIWC, LDA, word embedding), Image properties, including saturation, color theme, brightness, and color clarity, and transcript text of Audio. Most of the classification algorithms used are supervised learning(SL) algorithms like Support Vector Machines(SVM), Decision Tree(DT), Naïve Bayes(NB), Random Forest(RF), and Neural Networks(NN). The popular metrics considered are ERDE, Precision, Recall, F1-Score, Accuracy, $\boldsymbol{\mathrm{F}_{\text{latency}}}$. This paper presents a study of research for the early identification of depression in social media using machine learning (ML) techniques.

I. Introduction

A person's mental strength refers to his psychological, emotional, and social welfare, which influences how he or she thinks, feels, and acts. The World Health Organization (WHO) notes that mental illnesses are the primary cause of disability worldwide, affecting over 264 million adults, adolescents, and children. Depression, behavioral and mood disorders, suicidal and self-harming behavior, bipolar and borderline personality disorders, anxiety and panic attacks, phobias, schizophrenia, and others are amongst the most common mental disorders[1]. Depression is a psychological weakness with persistent sadness that interferes with one's ability to function and affects one's mood and emotions. Depression can vary in severity from slight to severe, that can last weeks, months, or even years. A person having depression experiences loss of interest in everyday tasks, insomnia, lack of concentration, significant weight gain or loss, sadness, and suicidal thoughts[2].