/Emotion_Detection_in-text

Web application to detect emotion in text

Primary LanguageJupyter Notebook

Emotion Classifier App

This is a Streamlit-based web application for detecting emotions in text using a pre-trained PySpark model. It allows users to input text, predicts the emotion associated with it, and provides a confidence score for the prediction. The app also provides monitoring capabilities to track page visits and analyze emotion classifier metrics.

Features

  • Emotion detection in text input.
  • Real-time prediction of emotions using a pre-trained PySpark model.
  • Visualization of prediction results using interactive charts.
  • Monitoring capabilities to track page visits and classifier metrics.
  • Database integration for storing page visit and prediction details.

Setup

  1. Clone the Repository:

    git clone https://github.com/yourusername/Emotion-Classifier-App.git
cd Emotion-Classifier-App

  2. Install Dependencies:

    pip install -r requirements.txt

  3. Set Up MySQL Database:

    • Ensure you have MySQL installed and running.
    • Create a database named DB.
    • Update database.py with your MySQL connection details.

  4. Run the Application:

    streamlit run app.py

Usage

  • Home: Allows users to input text for emotion detection. Predictions and confidence scores are displayed along with interactive charts showing prediction probabilities.
  • Monitor: Provides monitoring capabilities to track page visits and analyze emotion classifier metrics. Users can view page visit details, page metrics, and emotion classifier metrics.

Components

1. Backend

  • Spark Model Loading: Loads the pre-trained PySpark model for emotion detection.
  • Prediction Function: Defines a function to predict emotions using the loaded model.
  • Database Integration: Utilizes MySQL for storing page visit and prediction details.
  • Database Functions: Includes functions to create tables, add details, and view data from the database.

2. Frontend

  • Streamlit Application: Implements the web interface for the emotion classifier app.
  • User Input: Provides a text area for users to input text for emotion detection.
  • Prediction Display: Shows predicted emotions and confidence scores.
  • Interactive Charts: Visualizes prediction results using Altair and Plotly Express charts.
  • Monitoring: Enables users to monitor page visits and emotion classifier metrics.

Contributing

Contributions are welcome! Please feel free to open issues or submit pull requests.