Mental Health Analytics using Natural Language Processing

The Mental Health Analytics App is a Streamlit-based application designed to help mental health specialists in urban areas quickly understand their patients' mental health, particularly those in inaccessible rural areas. This tool bridges the gap between rural patients and therapists in urban centres by providing insights into patient moods, recurring thought patterns, and categorized concerns.

---

Features

1. Write Your Mood

• Allows users to share their thoughts and moods through text.
• Basic Version:
  • Performs sentiment analysis (Positive, Negative, Neutral).
  • Logistic Regression-based self-trained model with 80% accuracy.
• Pro Version:
  • Performs multi-emotion classification using a pre-trained DistilBERT model.
  • Visualizes emotion distribution over time.

2. Cognitive Corrections

• Uses Multimodal Naive Bayes classifier to classify recurring thoughts into 7 predefined categories such as:
  • Family, Finance, Relationships, Abuse, Health, Employment, Education.
• Tracks and visualizes recurring themes over time to highlight areas of concern.

3. Report Generation

• Generates a consolidated PDF report for therapists, including:
  • Mood trends and emotion distribution.
  • Categorized thought patterns.
  • Infographics for quick analysis.

---

Installation

Prerequisites

• Python 3.8+
• Git

Steps

1. Clone the repository:

   git clone https://github.com/ferozk0333/Mental-Health-Analytics-using-NLP.git
   cd Mental-Health-Analytics

2. Create Virtual Environment:

   python -m venv venv
   venv\Scripts\activate

3. Install Dependencies:

   pip install -r requirements.txt

4. Run app:

   streamlit run main.py

---

Future Enhancement Ideas

• Regional Language Support: Support local and regional languages to cater to a broader audience.