ITER Bhubaneswar Intelligent Chatbot

An intelligent chatbot for ITER Bhubaneswar that provides precise, contextual answers about the college using semantic search and natural language understanding.

🚀 Features

• Smart Query Understanding: Classifies user intent and extracts relevant context
• Precise Answers: Provides exact information without overwhelming details
• Course-Specific Responses: Automatically detects which course/branch you're asking about
• Semantic Search: Falls back to intelligent search when exact matches aren't found
• Multiple Interfaces: Both command-line and web interface available
• Self-Training: Learns from the JSON knowledge base to answer new questions

📁 Project Structure

/home/gaurav/Projects/Warp-1/
├── data_processor.py      # JSON data processing and search indexing
├── iter_chatbot.py       # Main chatbot logic with contextual understanding
├── app.py               # Flask web interface
├── templates/
│   └── index.html       # Web UI template
├── venv/                # Python virtual environment
└── README.md           # This file

🛠️ Installation

1. Clone/Navigate to project directory:

   cd /home/gaurav/Projects/Warp-1

2. Activate virtual environment:

   source venv/bin/activate

3. Dependencies are already installed:

   • scikit-learn (for TF-IDF and cosine similarity)
   • numpy (for numerical computations)
   • flask (for web interface)
   • Other standard libraries

💻 Usage

Command Line Interface

# Activate environment
source venv/bin/activate

# Run chatbot
python iter_chatbot.py

Web Interface

# Activate environment
source venv/bin/activate

# Start web server
python app.py

Then open http://localhost:5000 in your browser.

Testing Data Processor

# Test the data processing module
python data_processor.py

🤖 How It Works

1. Data Processing (data_processor.py)

• Loads JSON knowledge base
• Recursively extracts all searchable text
• Creates TF-IDF vectors for semantic search
• Provides specific info retrieval methods

2. Intelligent Understanding (iter_chatbot.py)

• Query Classification: Identifies intent (fees, placements, admissions, etc.)
• Course Extraction: Detects which course/branch is being asked about
• Context-Aware Responses: Provides precise answers based on detected intent
• Fallback Search: Uses semantic search when specific rules don't match

3. Smart Response Generation

• Exact Matches: "What is attendance requirement?" → "75% attendance is required"
• Course-Specific: "CSE fees" → "CSE fees: ₹137,500 per semester"
• Complex Queries: Handles multiple intents in one question
• Fallback: Semantic search for unmatched queries

🎯 Example Interactions

Query	Response
"What is the attendance requirement?"	"75% attendance is required"
"CSE fees per semester"	"CSE fees: ₹137,500 per semester"
"Average placement package"	"Average CTC: ₹5-9 LPA for engineering overall"
"AI and ML course details"	"B.Tech AI & ML: 4-year program. Cutting-edge AI, machine learning, data science curriculum."
"Which companies visit for placements?"	"Top recruiters include Amazon, Microsoft, TCS, Wipro, Infosys..."

🧠 Intelligence Features

Query Classification

The chatbot automatically understands:

• Fees queries: fee, cost, tuition, price
• Placement queries: placement, job, package, salary, ctc
• Admission queries: admission, entrance, eligibility, apply
• Course queries: course, program, degree, branch
• Facility queries: hostel, campus, lab, library, sports

Course Detection

Recognizes course mentions:

• CSE, Computer Science → CSE details
• AI, ML, Data Science → AI/ML program info
• ECE, Electronics → ECE program info
• And many more...

Context Awareness

• Maintains conversation history
• Provides concise, relevant answers
• Avoids information overload

🔧 Customization

Adding New Data

1. Update the JSON knowledge base at /home/gaurav/Downloads/iter_bhubaneswar_knowledge.json
2. The chatbot will automatically index new information
3. Add specific response patterns in get_specific_info() method if needed

Adding New Intent Categories

1. Update context_keywords dictionary in iter_chatbot.py
2. Add handling logic in _handle_single_intent() method
3. Test with relevant queries

Improving Search

• Adjust TF-IDF parameters in data_processor.py
• Modify similarity thresholds
• Add preprocessing rules for better text matching

🌐 Web Interface Features

• Responsive Design: Works on desktop and mobile
• Real-time Chat: Instant responses
• Example Queries: Click-to-try common questions
• Modern UI: Clean, professional interface
• Error Handling: Graceful error messages

📊 Technical Details

• Search Engine: TF-IDF with cosine similarity
• Text Processing: Recursive JSON traversal
• Query Understanding: Pattern matching + semantic search
• Response Generation: Rule-based + retrieval hybrid
• Web Framework: Flask
• Frontend: Pure HTML/CSS/JavaScript

🎯 Key Benefits

1. Precise Answers: No information overload - just what you asked for
2. Context Aware: Understands what course/topic you're interested in
3. Natural Language: Ask questions naturally, no specific format required
4. Fast Response: Instant answers with efficient search
5. Always Learning: Automatically incorporates new information from JSON updates

🚀 Future Enhancements

• Add more sophisticated NLP models
• Implement conversation memory
• Add voice interface
• Create admin panel for knowledge base management
• Add analytics and usage tracking
• Multi-language support

📝 License

This project is created for ITER Bhubaneswar and is intended for educational and informational purposes.

---

Contact: For questions about this chatbot, you can ask it directly! 😊

The chatbot knows about:

• Course fees and details
• Placement statistics
• Admission procedures
• Campus facilities
• Contact information
• And much more about ITER Bhubaneswar!