During the week 3 and week 4, I focused on the Assessment 2 Project proposal and plan.

Research significance

The World Health Organization reported 970 million mental disorders globally, with anxiety and depressive disorders being the most common. COVID-19 has increased these numbers, necessitating innovative early identification and treatment methods. Social media provides user-generated data for mental health assessments. This project aims to create a dashboard using machine learning, natural language processing, and R programming languages to analyze social media text to identify mental health issues, improving the accuracy and reliability of mental health predictions.

Aim

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.

Objectives

• Develop a Comprehensive Dashboard: Create a user-friendly dashboard using the R language that integrates NLP and ML techniques to analyze social media text for mental health indicators.
• Provide an idea for the user about most popular words in social media text according to mental health disorder.
• Text Data Preprocessing: Collect and preprocess a diverse range of social media texts to build a robust dataset for analysis.
• Feature Extraction and Model Training: Utilize NLP techniques to extract relevant features from the text data and train ML models to predict mental health conditions.
• Real-Time Analysis and Prediction: Enable real-time analysis and prediction capabilities within the dashboard, allowing users to input their social media text and receive immediate feedback on potential mental health issues.
•  Validation and Testing: Conduct thorough validation and testing of the dashboard with real-world data to ensure its effectiveness and reliability.

I did further investigation on research papers to address following research questions.

Main Question:
How do we detect mental disorders and intervene early using AI technologies?
Sub question:

1. How can NLP be used for the prediction and analysis of mental health conditions from text, to develop the dashboard?
2. What is the most suited classification method to classify text and documents in machine learning specially in supervised learning to develop the dashboard?
3. How can user understand that the terms that are indicative of mental disorders?

I found interesting research articles related to my topic. Some of them are Quantitative and others are qualitative. One article is a survey article. Kennard et al. (2023), their study examined suicidal thoughts and behaviors in youth seeking mental health treatment in Texas using 1000 participants. By reading research articles, I was able to write a summary of my Preliminary Literature Review using 6 articles.

Next I focused on the research methodology of my project. This research employs a quantitative approach, with text analysis serving as the principal methodology. The research stages include 8 stages as below.

Research Methodology

The research will be undertaken through the following steps:

1. Data Collection: Collecting social media texts relevant to mental health conditions.
2. Data Preprocessing: Cleaning and preprocessing the textual data for analysis.
3. Model Development: Developing NLP and ML models to classify and predict mental health conditions from text.
4. Dashboard Development: Creating a user-friendly dashboard in R for inputting social media texts and predicting mental health conditions.
5. Model Evaluation: Evaluating the accuracy and effectiveness of the ML models using appropriate metrics.
6. Literature Review: Conducting a comprehensive review of existing research on NLP and ML in mental health analysis.
7. User Testing: Testing the dashboard with users to ensure usability and accuracy.
8. Documentation: Creating user guidelines and a future plan for the dashboard.

Next, I will focus on plan of my project using a Work Breakdown Structure (WBS) and a Gantt Chart.

Reference

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/10.1111/sltb.12980

World Health Organization. (2024). Mental disorders https://www.who.int/news-room/fact-sheets/detail/mental-disorders