Tourism Knowledge Graph QA 🌍🗺️

A question-answering system about tourist places and countries using a Neo4j knowledge graph powered by Wikidata data.

Screenshots:

1. Gradio App Interface:

2. Neo4j Graph Database:

Features

• 🏞️ Small curated dataset of 500+ touristic attractions and 600+ countries
• 🔗 Relationship mapping (LOCATED_IN) between attractions and countries
• 💬 Natural language query interface using Gemma-2B LLM
• 📊 Extended properties including geographical metrics and demographic data

Dataset Structure

Nodes:

• TouristicAttraction: Name, type, dimensions (length/width/area)
• Country: Continent, capital, language, population

Relationship:

• (Attraction)-[LOCATED_IN]->(Country)

🛠️ Tools & Libraries

• Graph Database: Neo4j
• LLM Integration: LangChain + Groq API (Gemma-2B)
• Data Sources: Wikidata SPARQL API
• UI Framework: Gradio
• Core Libraries:
  • langchain_neo4j - Neo4j graph integration
  • langchain_groq - LLM inference
  • python-dotenv - Environment management
  • SPARQLWrapper - Wikidata query interface

🚀 Installation & Setup

Create environment

conda create -n kg python=3.11 -y
conda activate kg-env

Install dependencies

pip install -r requirements.txt

Configure environment

echo "NEO4J_URI='uri'
NEO4J_USERNAME='neo4j'
NEO4J_PASSWORD='your_password_here'
GROQ_API_KEY='your_groq_key_here'" > .env

Run application

python app.py

Access Interface

Visit http://127.0.0.1:7860/ in your browser