Depression Detection Using Sentiment Analysis of Social Media Data

Abstract

Depression is one of the major causes of increased number of cases of mental illness and suicides all over the globe now a days. A person suffering from anxiety, sadness, depression or suicidal thoughts finds it easier to express his emotions on social media platforms. Thus, the messages or content shared by a person on social media platforms is the best way to detect the mental condition of a person by analysing these messages. Today’s situation of pandemic covid19 also increased the cases of depression. In this paper we have used Natural Language processing techniques and deep learning methods to create a model to predict such mental conditions like depression. The model created using LSTM-CNN gives better accuracy of 97% when compared to the other base models of logistic regression, Naïve Bayes, Random Forest and Decision Tree.