Mental Health Prediction Using Natural language Processing and Machine Learning Approach

Research Supervisor
Journal Title
Journal Title
Journal ISSN
Volume Title
Due to the rise in mental health problems and the demand for efficient medical care, machine learning has been looked into as a potential solution for mental health disorders. This research presents analysis on how mental health disorders can be detected through Natural Language Processing (NLP).It deals with the capacity of computers to comprehend textual information in a manner similar to that of humans. The ability of NLP to process large amounts of unstructured data effectively and the way NLP analyzes text data more effectively than traditional methods, makes it well-suited for predicting mental health conditions. Digital text is subjected to sentiment analysis in an effort to extract human emotions. With the help of NLP sentiment analysis, we could analyze human emotions through the data provided on social networks and can come to conclusions that will help people who are suffering from mental illness and thus help in correct treatment. Currently our project aims to detect 5 types of mental illness such as Anxiety, Depression, PTSD, Social-anxiety and Suicidal thoughts. In this study we have used five machine learning techniques, one deep learning model and assessed their accuracy in identifying mental health issues using several accuracy criteria. The ml techniques are Random Forest, Linear SVC, Naive bayes, Logistic Regression, XGB and the Deep learning model is LSTM. We have compared these techniques and implemented them and also obtained the highest accuracy with Logistic regression technique which is 79.3%.
TECHNOLOGY::Electrical engineering, electronics and photonics::Electrical engineering
Department Name
North South University
Printed Thesis