Emotion & Risk Classification Chatbot Version 1.5

Overview

The Emotion & Risk Classification Chatbot is a real-time NLP application designed to analyze emotional tone and contextual relapse-risk language using machine learning and sentence embeddings.

The project was built to explore how natural language processing can be applied to emotionally complex text classification scenarios while creating an accessible and interactive user experience through streamlit deployment.

The application classifies text into multiple emotional and contextual categories, including:

• Positive
• Neutral
• Negative
• High Risk

Project Goals

This project was created to:

• Explore practical NLP and machine learning workflows
• Build a deployable AI-powered web application
• Experiment with emotional language classification
• Practice model training, prediction pipelines, and deployment
• Create a portfolio-ready end-to-end machine learning project

Technologies Used

Programming & Frameworks

• Python
• Streamlit

Machine Learning & NLP

• scikit-learn
• Sentence Transformers
• all-MiniLM-L6-v2 embeddings

Data & Model Tools

• pandas
• NumPy
• joblib

Deployment & Development

• GitHub
• Streamlit
• Community Cloud

Features

• Real-Time emotional text analysis
• Context-aware risk classification
• Sentence embedding processing
• Interactive Streamlit web interface
• Probability confidence scoring
• Dynamic chatbot-style responses
• Cloud deployment for public access

Deployment Process

The project involved:

• Dataset preperation and labeling
• NLP preprocessing workflows
• Sentence embedding generation
• Model training and evalutation
• Prediction pipeline integration
• UI refinement using Streamlit
• GitHub version control development

Through development, the focus remained on combining technical experimentation with meaningful human-centered language analysis.

Challenge & Learning

During development, I learned:

• How to structure machine learning projects for deployment
• GitHub workflow fundamentals
• Streamlit deployment and debugging
• NLP embedding workflows
• Model prediciton handling and probability scoring
• Front-end refinement for interactive AI applications

This project significantly expanded my practical understanding of AI application development and deployment pipelines.

Future Improvements

Planned future enhancements may include:

• Expanded dataset training
• Improved classification accuracy
• Sentiment trend visualization
• Conversation history tracking
• Advanced contextual analysis
• Enhanced UI/UX design
• Additional emotional classification categories

Creator Note

This project represents a transition into applied AI and NLP development through self-directed learning, experimentation, and hands-on project building.

• GitHub
• LinkedIn
• Instagram