Bayesian Knowledge Extraction System

A Python-based Artificial Intelligence project that extracts structured knowledge from text and represents relationships using a Bayesian Network. The system processes natural language data, builds probabilistic relationships, and visualizes connections using Neo4j graph database.

Project Overview

Understanding and structuring knowledge from unstructured text is a key challenge in Artificial Intelligence. This project demonstrates how Natural Language Processing (NLP), Bayesian Networks, and Graph Databases can work together to transform raw text into structured and visual knowledge.

The system performs the following steps:

1. Parses natural language text
2. Extracts entities and relationships
3. Populates structured knowledge
4. Builds probabilistic dependencies using Bayesian Networks
5. Visualizes relationships using Neo4j graph database

This project demonstrates a simplified AI knowledge representation pipeline.

---

Features

• Natural language text parsing
• Entity and relationship extraction
• Knowledge base population
• Bayesian Network representation
• Graph visualization using Neo4j
• Modular and extendable Python implementation

---

Project Structure

Bayesian-Network-Knowledge-System
│
├── bayesianNetwork.py        # Creates and manages the Bayesian Network
├── parsingWithSpacy.py      # NLP parsing using spaCy
├── populateData.py          # Populates extracted knowledge
├── fileParsing.py           # Reads and processes input text
├── knowledge.txt            # Sample knowledge dataset
├── test_connection.py       # Neo4j connection test
└── README.md

---

Technologies Used

• Python
• spaCy (Natural Language Processing)
• Bayesian Networks
• Neo4j Graph Database
• Artificial Intelligence Concepts
• Knowledge Representation

---

Why Neo4j?

Neo4j is used in this project to visually represent extracted knowledge as a graph structure.

Instead of storing relationships in traditional tables, Neo4j represents them as:

• Nodes → Entities
• Relationships → Connections between entities

This makes it easier to explore and analyze knowledge structures created from text data.

Example representation:

Entity A  ----->  Entity B
    |               |
  related_to     influences

Using Neo4j allows the system to display these relationships interactively, making the knowledge network easier to understand.

---

Installation

Clone the repository:

git clone https://github.com/yourusername/Bayesian-Network-Knowledge-System.git
cd Bayesian-Network-Knowledge-System

Install dependencies:

pip install spacy
python -m spacy download en_core_web_sm

Install Neo4j and start the Neo4j database before running the project.

---

How to Run

Run the NLP parser:

python parsingWithSpacy.py

Populate extracted data:

python populateData.py

Test the Neo4j connection:

python test_connection.py

---

Learning Objectives

This project demonstrates important Artificial Intelligence concepts including:

• Knowledge extraction from text
• Natural Language Processing pipelines
• Probabilistic reasoning
• Bayesian Networks
• Graph-based knowledge visualization
• Integration of AI with graph databases

---

Future Improvements

Possible improvements include:

• Visualization of full Bayesian network dependencies
• Support for larger datasets
• Integration with knowledge graphs
• Advanced NLP pipelines
• Interactive dashboard for graph exploration

---

Author

Moiz Malik BS Artificial Intelligence Student

---

License

This project is open-source and available for educational purposes.