danb-neo4j / gds-guide

Practitioner's Guide to Getting Started with Neo4j GDS

Practitioner's Guide to Getting Started with Neo4j GDS

Author: Daniel Bukowski, Graph Data Science Customer Success Architect, Neo4j

Last Updated: 27 June 2023

Description

This document is a practical, hands-on guide for data scientists encountering Neo4j and Graph Data Science for the first time. The intent is to provide a critical path roadmap for what to learn and in what order to learn it, focusing on practical, hands-on-keyboard resources (most of which already exist) that can help the reader begin using Neo4j and GDS to produce results as quickly as possible. As this document evolves it will link to more resources, especially industry-specific and use case-specific resources.

Note: This repository is an unofficial guide intended to complement the official Neo4j documentation and official Neo4j Support.

Audience

Anyone starting to use the Neo4j GDS library for the first time, with an emphasis on data scientists.

For the initial iteration, the document will make the following assumptions about the reader:

• Working in a data analyst or data science role
• Understands of data science concepts
• Comfortable with Jupyter Notebooks and has a basic-to-intermediate knowledge of data science Python (pandas, matplotlib/seaborn, scikit-learn, etc…)
• New to graphs, Neo4j, and graph data science

How to Use this Repo

• The intent is to define the critical path from an initial introduction to Neo4j to using the database and library to create value.
• Not everyone will start from the same place or with the same background knowledge, so the intent is also to enable users to jump to the specific topics where they have questions or need assistance.
• You do not need to review all of the resources in each section. There may be documentation, book chapters, blog articles, videos, or notebooks, so leverage those that best match your learning style and immediate need.

Table of Contents

Neo4j Database Resources:

• Graph Database and Neo4j Overview
• Neo4j Memory Configuration
• Cypher
• APOC Library
• Data Loading

EDA and Graph EDA

Graph Data Modeling

• Graph Data Modeling Overview
• Temporal Graph Data Modeling

Neo4j Graph Data Science Resources

• General Graph Data Science
• Graph Data Science Algorithms
• Node Embedding Algorithms

Usecase-Specific Resources

• Fraud and Anti-Money Laundering
• Customer and Patient Journey
• Recommendation Systems
• Machine Learning Features
• Logistics
• Healthcare

Graphs and Large Language Models (LLMs)