Project Brief

Project Title

A dashboard to predict mental health disorders using NLP and Machine Learning by analysing social media texts.

Abstract

Mental health issues are on the rise due to a combination of societal, technological, environmental, and individual factors. Conditions like depression, anxiety, and stress-related disorders are becoming more common, affecting people of all ages and backgrounds. The rising prevalence of mental health issues has necessitated innovative approaches to early detection and intervention. Social media, as a significant part of daily communication, offers a vast amount of user-generated data that can be analysed for mental health indicators. This project aims to develop a dashboard using R language, Natural Language Processing (NLP) and Machine Learning (ML) techniques to analyse social media texts, identifying mental health conditions. The proposed framework will predict mental health issues by analysing textual data, providing a valuable software tool to everyone. Mainly, people can identify the mental disorder by entering social media text using the proposed dashboard. This project addresses the challenge of leveraging social media data for mental health insights, aiming to contribute to more effective mental health management and support systems.

Provide a brief description as to how does the selected project align with your course or specialisation?

This project aligns with my specialisation in software development within my Master’s degree in Professional IT by integrating advanced software engineering principles with cutting-edge NLP and ML techniques. It involves developing a sophisticated software tool using R programming language, aligning with the core competencies of software development, and addressing a significant societal issue, demonstrating the practical application of my academic training.

Background/Description: 

A number of causes, including cultural, technical, environmental, and personal factors, are contributing to the growth in mental health disorders. Increased awareness and diagnosis of mental health issues, increased social media use, economic pressures, social isolation, workplace stress, environmental factors like climate change, natural disasters, and pandemics, chronic health conditions, substance abuse, rapid social changes, and family dynamics and childhood adversities contribute to higher reported rates. Economic instability, job insecurity, and high living costs also contribute to increased stress, anxiety, and depression. Chronic health conditions, substance abuse, rapid social changes, and family dynamics can exacerbate mental health issues.

Mental health issues are increasingly prevalent, with social media providing a unique window into users’ psychological states through their posts and interactions. The analysis of this data can offer insights into mental health trends and individual conditions, but it requires advanced techniques to process and interpret the vast and unstructured nature of social media text.

Project Aim/Objectives: 

This project aims to address the gap in proactive mental health monitoring by leveraging social media data, thus enabling earlier detection and potentially better outcomes by developing a software solution for predicting mental health conditions from social media text using NLP and Machine Learning.

Project Problem Domain:

Mental health is a critical area of concern globally, with many individuals not receiving timely diagnosis or intervention. Traditional methods of mental health assessment are often reactive and resource-intensive.

Following research questions will be investigated

  • How NLP can be used for this prediction and analysis of text?
  • What is the most suited classification method to classify text and documents in machine learning specially in supervised learning?
  • What are the advantages of machine learning over traditional methods?

Deliverables/Outcomes:

  • A dashboard implementing NLP and ML model to predict mental health conditions using social media texts.
  • A comprehensive literature review on the use of NLP and ML in mental health analysis.
  • Present the accuracy of ML model.
  • User guidelines on how to use this dashboard to predict mental disorder.
  • Future plan for this dashboard.

Resources:

Ananthakrishnan, G., Jayaraman, A. K., Trueman, T. E., Mitra, S., Abinesh, A. K., & Murugappan, A. (2022). Suicidal intention detection in tweets using BERT-based transformers. In 2022 International Conference on Computing, Communication, and Intelligent Systems (ICCCIS) (pp. 322-327). IEEE. https://doi.org/10.1109/ICCCIS56430.2022.10037677

Chadha, A., & Kaushik, B. (2022). A hybrid deep learning model using grid search and cross-validation for effective classification and prediction of suicidal ideation from social network data. New generation computing, 40(4), 889-914. https://link.springer.com/article/10.1007/s00354-022-00191-1

Chatterjee, M., Samanta, P., Kumar, P., & Sarkar, D. (2022). Suicide ideation detection using multiple feature analysis from Twitter data. In 2022 IEEE Delhi Section Conference (DELCON) (pp. 1-6). IEEE. https://doi.org/10.1109/DELCON54057.2022.9753295

Kennard, B. D., Hughes, J. L., Minhajuddin, A., Slater, H., Blader, J. C., Mayes, T. L., & Trivedi, M. H. (2023). Suicidal thoughts and behaviors in youth seeking mental health treatment in Texas: Youth Depression and Suicide Network research registry. Suicide and Life‐Threatening Behavior, 53(5), 748-763. https://doi.org/10.1109/TLA.2023.10172137

Santos, W., Yoon, S., & Paraboni, I. (2023). Mental health prediction from social media text using mixture of experts. IEEE Latin America Transactions, 21(6), 723-729. 10.1109/TLA.2023.10172137