EmotiAI – AI-Powered Text Emotion Detection

EmotiAI is an AI-powered text emotion detection system that analyzes English text and identifies the emotion expressed in the input.

The system uses the DAIR.AI Emotion dataset and a machine-learning pipeline based on TF-IDF feature extraction and a Linear Support Vector Machine (Linear SVM) classifier. The text is cleaned, converted into numerical TF-IDF features, and then classified into one of six emotion categories: sadness, joy, love, anger, fear, or surprise.

The application provides an interactive web interface where users can enter text and receive an emotion prediction. Along with the predicted emotion, EmotiAI provides a visual representation of the model's decision scores across all six emotion classes.

The system also maintains prediction history and provides application-level statistics such as the total number of predictions, emotion distribution, and most frequently detected emotion.

A dedicated contact system allows visitors to submit their name, mobile number, email address, and message. Contact submissions are stored in the application's database and can be managed through the Django administration panel.

The website includes configurable social and contact links for Instagram, LinkedIn, and email, which can be managed through the administration panel.

EmotiAI is designed for English text emotion classification and supports six emotion categories.

1. Visit the EmotiAI website.
2. Navigate to the Model page.
3. Enter English text into the emotion analysis text box.
4. Click "Analyze Emotion".
5. The application sends the text to the Django backend.
6. The text is cleaned and transformed using the trained TF-IDF vectorizer.
7. The Linear SVM model predicts one of six emotions.
8. The detected emotion is displayed with an explanation.
9. A visual decision-score chart displays the model's scores for all six emotions.
10. The prediction is stored in the database.
11. Recent predictions can be viewed in the prediction history.
12. Application statistics are updated automatically.
13. Visitors can use the Contact page to submit their details and message.
14. Administrators can manage predictions, contact messages, and footer social/contact links through Django Admin.

Python 3.13 or compatible Python 3.x version
Django 5.x
scikit-learn
joblib
SQLite
Modern web browser
Virtual environment recommended