Unibot University Chatbot System

📋 Overview

UniBot is an intelligent, AI-powered chatbot system designed to provide real-time, accurate, and context-aware assistance for university-related queries. Built using Meta's LLaMA model with Retrieval-Augmented Generation (RAG) and structured function calling, UniBot addresses the limitations of traditional chatbots by offering:

• High Accuracy: RAG-enhanced responses grounded in official university documents
• Low Latency: Ultra-fast inference powered by Groq hardware acceleration
• Dynamic Information Retrieval: Real-time access to course catalogs, faculty directories, and administrative data
• Contextual Understanding: Multi-turn conversation support with memory retention
• Extensible Architecture: Modular design for easy integration of new services

✨ Key Features

🎓 Academic Support

• Course Information: Detailed course descriptions, prerequisites, credits, and schedules
• Faculty Lookup: Find professors by name or department with contact information and office hours
• Subject Search: Browse courses by department with comprehensive details

💰 Administrative Assistance

• Fee Information: Tuition costs, additional fees, and payment plan details
• Policy Search: RAG-powered semantic search across university policies
• FAQ System: Quick answers to common student questions

🤖 Advanced AI Capabilities

• Function Calling: Structured tool execution for precise data retrieval
• RAG Pipeline: Document-grounded responses to prevent hallucinations
• Hybrid Fallback: Automatic switching between retrieval and tool-based responses
• Multi-turn Conversations: Contextual memory for natural dialogue flow

🎨 User Experience

• Dark/Light Mode: Professional theme toggle for comfortable viewing
• Responsive Design: Optimized for desktop and mobile devices
• Interactive Chat Interface: Real-time responses with visual feedback
• Quick Actions: Sidebar buttons for common queries

🏗️ System Architecture

┌─────────────────┐
│  User Interface │ (Streamlit)
└────────┬────────┘
         │
    ┌────▼─────┐
    │  Query   │
    │  Parser  │
    └────┬─────┘
         │
    ┌────▼──────────┐
    │ Tool Resolver │
    └───┬───────┬───┘
        │       │
   ┌────▼───┐ ┌─▼────────┐
   │Function│ │   RAG    │
   │Calling │ │ Pipeline │
   └────┬───┘ └─┬────────┘
        │       │
    ┌───▼───────▼───┐
    │  LLaMA Model  │ (Groq-accelerated)
    │   Generator   │
    └───────┬───────┘
            │
    ┌───────▼────────┐
    │    Response    │
    │   Formatter    │
    └────────────────┘

🚀 Installation

Prerequisites

• Python 3.10 or higher
• Virtual environment (recommended)
• Groq API key (for LLaMA access)

Setup Instructions

1. Clone the repository

   git clone <repository-url>
   cd chatbot
   

2. Create and activate virtual environment

   # Windows
   python -m venv venv
   .\venv\Scripts\activate
   
   # Linux/Mac
   python3 -m venv venv
   source venv/bin/activate
   

3. Install dependencies

   pip install -r requirements.txt
   

4. Set up environment variables (Optional)

   # Create .env file
   echo "GROQ_API_KEY=your_api_key_here" > .env
   

   Or set as environment variable:

   # Windows
   set GROQ_API_KEY=your_api_key_here
   
   # Linux/Mac
   export GROQ_API_KEY=your_api_key_here
   

5. Run the application

   streamlit run app.py
   

6. Access the chatbot

   • Open your browser and navigate to http://localhost:8501

📦 Technology Stack

| Component | Technology | Purpose | |-----------|-----------|---------| | LLM | Meta LLaMA 3.1 (8B) | Natural language understanding and generation | | Acceleration | Groq Hardware | Ultra-low latency inference (~0.42s response time) | | Framework | LangChain | Agent orchestration and tool integration | | Embeddings | MiniLM-L6-v2 | Semantic document encoding for RAG | | Vector Store | FAISS | Fast similarity search for document retrieval | | UI | Streamlit | Interactive web interface | | Language | Python 3.10+ | Core implementation |

📊 Performance Metrics

Based on evaluation with 10 participants (students and staff):

| Metric | Result | |--------|--------| | Response Accuracy | 92% (with RAG) vs 78% (closed-book) | | Average Latency | 0.42s (Groq) vs 1.35s (standard hardware) | | User Satisfaction | 90% found system easy to use | | Contextual Relevance | 80% of responses rated as helpful | | Hallucination Reduction | 25% improvement with RAG |

🎯 Usage Examples

Example 1: Course Information

User: "What are the prerequisites for CS201?"
UniBot: "CS201 - Data Structures and Algorithms requires CS101 as a prerequisite. 
         It's a 4-credit course offered in both Fall and Spring semesters."

Example 2: Faculty Lookup

User: "Who teaches in the Computer Science department?"
UniBot: "Here are the teachers in Computer Science:
         - Dr. Sarah Johnson, Professor (sarah.johnson@university.edu)
         - Prof. Michael Chen, Associate Professor (michael.chen@university.edu)"

Example 3: Policy Search

User: "What is the grading policy?"
UniBot: "According to university policy, the grading system uses:
         A (90-100%): 4.0 grade points
         B (80-89%): 3.0 grade points
         [Full policy details provided...]"

🛠️ Configuration

Customizing Data Sources

1. Update Faculty Information: Edit teachers.json
2. Modify Course Catalog: Edit subjects.json
3. Adjust Fee Structure: Edit fees.csv
4. Update FAQs: Edit faq.txt
5. Revise Policies: Edit policies.txt

Adjusting Model Parameters

In app.py, modify the LLM initialization:

ChatGroq(
    api_key=api_key,
    model="llama-3.1-8b-instant",  # Model selection
    temperature=0.7                 # Creativity (0.0-1.0)
)

🔧 Troubleshooting

Common Issues

Port Already in Use

streamlit run app.py --server.port=8502

Module Import Errors

pip install --upgrade -r requirements.txt

Slow Response Times

• Check internet connection (Groq API requires network access)
• Verify Groq API key is valid
• Consider reducing document corpus size

Outdated Policy Retrieval

• Update document files (faq.txt, policies.txt)
• Restart the application to reload vector embeddings

📈 Future Enhancements

• [ ] Multilingual support
• [ ] Voice interaction capabilities
• [ ] Integration with university LMS (Learning Management System)
• [ ] Automated policy document updates
• [ ] Enhanced emotion recognition
• [ ] Mobile application
• [ ] Analytics dashboard for administrators

🤝 Contributing

Contributions are welcome! Please follow these steps:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/AmazingFeature)
3. Commit your changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

👥 Authors

• Development Team - Initial work and implementation
• Based on research documented in "Development of a University Chatbot using Retrieval-Augmented Generation (RAG) with Custom Vector-Based Knowledge Base"

🙏 Acknowledgments

• Meta AI for the LLaMA model
• Groq for hardware acceleration platform
• LangChain community for the agent framework
• Streamlit for the UI framework
• All contributors and testers

📞 Support

For questions, issues, or feature requests:

• Open an issue on GitHub
• Contact: support@university.edu
• Documentation: Wiki

Note: This is a prototype system designed for educational purposes. For production deployment, ensure compliance with data privacy regulations (GDPR, FERPA) and implement appropriate security measures.