Description
For most modern-day data, graph data models are proving to be advantageous since they facilitate a diverse range of data analyses. This has spiked the interest and usage of graph databases, especially Neo4j. We study Neo4j and cypher along with various plugins that augment database capabilities in terms of data types or facilitate applications in data science and machine learning using plugins like graph data science (GDS).A significant portion of the book is focused on discussing the structure and usage of graph algorithms. Readers will gain insights into well-known algorithms like shortest path, PageRank, or Label Propagation among others, and how one can apply these algorithms in real-world scenarios within a Neo4j graph.Once readers become acquainted with the various algorithms applicable to graph analysis, we transition to data science problems. Here, we explore how a graph's structure and algorithms can enhance predictive modeling, prediction of connections in the graph, etc. In conclusion, we demonstrate that beyond its prowess in data analysis, Neo4j can be tweaked in a production setup to handle large data sets and queries at scale, allowing more complex and sophisticated analyses to come to life.
Key Features
? Utilizing graphs to improve search and recommendations on graph data models.
? Understand GDS and Neo4j graph algorithms including cluster detection, link prediction, and centrality.
? Complex problem-solving for predicting connections, application in ML pipelines and GNNs using graphs.
What you will learn
? Understand Neo4j graphs and how to effectively query them with cypher.
? Learn to employ graphs for effective search and recommendations around graph data.
? Work with graph algorithms to solve problems like finding paths, centrality metrics, and detection of communities and clusters.
? Explore Neo4j's GDS library through practical examples.
? Integrate machine learning with Neo4j graphs, covering data prep, feature extraction, and model training.
Who this book is for
The book is intended to serve as a reference for data scientists, business analysts, graph enthusiasts, and database developers and administrators who work or intend to work on extracting critical insights from graph-based data stores.
Table of Contents
1. Data Representation as Graphs – Introducing Neo4j
2. Processing Graphs with Cypher Queries
3. A Peek into Recommendation Engines and Knowledge Graphs
4. Effective Graph Traversal and the GDS Library
5. Centrality Metrics, PageRank, and Fraud Detection
6. Understanding Similarity and Cluster Analysis Algorithms
7. Applications of Graphs to Machine Learning
8. Link Prediction with Neo4j
9. Embedding, Neural Nets, and LLMs with Graphs
10. Profiling, Optimizing, and running Neo4j and GDS in Production