Ahmad Tawil

IBM

Developing AI Applications with Python and Flask

Flask + Watson NLP emotion-detection capstone — classifies anger, disgust, fear, joy, and sadness from customer feedback text.

Completed 2025-09-16

Verify credentialCoursework repository

🎭 Emotion Detection Web Application #

An AI-powered web app that detects emotions — anger, disgust, fear, joy, sadness — from customer feedback text.
Developed with Flask and Watson NLP (Embeddable Libraries) as the final project of the IBM Developing AI Applications with Python and Flask course.

📌 Overview #

This project extends sentiment analysis into fine-grained emotion detection.
Instead of just positive/negative polarity, it identifies which emotion dominates a piece of text.

✔️ Input: Free-form text from the user
✔️ Processing: Watson NLP EmotionPredict model
✔️ Output: JSON with emotion scores + a formatted natural language response

📊 Example #

Input:
I love my life.

JSON Output:

{
  "anger": 0.0062,
  "disgust": 0.0026,
  "fear": 0.0093,
  "joy": 0.9680,
  "sadness": 0.0497,
  "dominant_emotion": "joy"
}

Browser Display:

For the given statement, the system response is 
'anger': 0.0062, 'disgust': 0.0026, 'fear': 0.0093, 
'joy': 0.9680 and 'sadness': 0.0497. 
The dominant emotion is joy.

🛠️ Tech Stack #

  • Python 3.10+
  • Flask — backend web framework
  • Requests — HTTP client for Watson NLP calls
  • Watson NLP Embeddable Library — emotion detection models
  • HTML/CSS/JavaScript — provided templates and scripts
  • PyLint / Unittest — static code analysis & validation

📂 Project Structure #

EmotionDetection/        # Packaged emotion detection module
    __init__.py
    emotion_detection.py
static/                  # Provided JavaScript file
templates/               # Provided index.html
server.py                # Flask web server
test_emotion_detection.py
requirements.txt
README.md

▶️ Running Locally #

  1. Clone the repository

    git clone https://github.com/<your-username>/emotion-detector-app.git
cd emotion-detector-app

  2. Create and activate a virtual environment

    python -m venv .venv
source .venv/bin/activate   # Mac/Linux
.venv\Scripts\activate      # Windows

  3. Install dependencies

    pip install -r requirements.txt

  4. Run the Flask server

    python server.py

  5. Access the app

    http://127.0.0.1:5000

✅ Features #

  • Detects five core emotions using Watson NLP.
  • Clean Flask architecture for easy extension.
  • Error handling for invalid/blank input.
  • Unit tests validating joy, anger, disgust, sadness, fear.
  • PyLint 10/10 compliance (Task 8).

🎓 Certification #

This project was developed as part of the IBM Developing AI Applications with Python and Flask course.

📜 Certificate: Verify on Coursera

🚀 Future Work #

  • 🌐 Add multi-language emotion detection.
  • ☁️ Deploy on cloud platforms (Render, IBM Cloud, Vercel).
  • 📈 Visualize emotions using interactive charts.

👨‍💻 Author #

Ahmad Tawil
Software Engineering Student @ Braude College
Passionate about AI, Machine Learning, and Full-Stack Development

Developing AI Applications with Python and Flask — Ahmad Tawil