If one follows software development, it's almost impossible to miss the buzz about NoSQL databases. Graph databases get mentioned along with the more familiar document and column-family NoSQL solutions. However, they seem to be frequently relegated to the periphery of the discussion given they are thought to address very specific problem domains - namely recommendation engines and social network analysis. More people, myself included, are now realizing that many real-world entities can be modeled as networks of data and that valuable insights can be realized if they could be understood better. Historically, modeling such networks via traditional relational databases has always been difficult. With all that said, graph databases are now being given more attention.I think few would disagree that whenever graph databases are mentioned, Neo4j gets mentioned first as an actual example. Therefore, I read Learning Neo4j, by Rik Van Bruggen and published by Packt Publishing, with great interest. I can say this is the first complete book I have read about Neo4j so I have not been influenced by any prior resources.The About the Author section states that Rik Van Bruggen is a Regional Territory Manager for Neo Technology - the company behind Neo4j. I expected a fair amount of product-specific evangelism when I first started to read the book. I was pleased to see that Rik kept that to a minimum and did so in a very professional manner. The focus stays on conveying the basic concepts and I never felt feel like I was reading a sales pitch from a vendor whitepaper.Rather than just provide a very high-level review, interested readers may find the following chapter-level detail to be more informative.Chapter 1 - (Graphs and Graph Theory - An Introduction) - The author sets the foundation by explaining graph theory in an understandable way. I really don't see how this could have been done better. Also included was some history going back to the famous mathematician Euhler who is considered by most to be the father of graph theory. By graphs, we mean nodes and edges- not graphics. I thought the illustration from Euhler's original 1736 paper was a nice touch. The author goes on to discuss the basic definitions of graphs and how they apply to real-world problems. The final message of this chapter: "Graphs Are Everywhere."Chapter 2 (Graph Databases - Overview) - This chapter discussed how the traditional relational database systems suffer when attempting to model data organized as networks. The author discusses the family of NoSQL databases and how they work such as key-value, column-family, and document stores along with graph-oriented databases. The author goes on to describe a key concept called the Property Graph model along with a description of what graph databases are good at and not so good at. The amount of territory covered in this chapter and the way it was written was nicely done.Chapter 3 (Getting Started with Neo4j) - In this chapter, one gets into Neo4j. Probably the biggest surprise for me was reading that Neo4j is an ACID-compliant database. The author spends some time discussing the differences between OLTP and OLAP databases along with the declaration that Neo4j falls into the OLTP category. Also, the reader gets their first introduction to Cypher - the query language for performing DDL and DML operations on Neo4j databases. Finally, the author provides installation steps for Windows, Mac, and Linux operating systems. The author also makes a brief mention of Neo4j in a cloud environment. I was very happy to read about a cloud-based "graph as a service" solution called GrapheneDB. I plan on checking into that further.Chapter 4 (Modeling Data for Neo4j) - I regard this chapter as a core part of this book and also the most difficult to internalize. In this chapter, the author goes into detail on how to model real world entities into the graph database structure. I feel the author makes a good case explaining the limitations and complexities associated with modeling the real-world through the use of entity-relationship (ER) diagrams used for relational database design. This is contrasted against the flexibility of a graph database. Hence, the title of one section: "A graph model - a simple, high-fidelity model of reality". I would say the critical point made in this chapter is the questions you are trying to answer drive the design of the data model. In other words, the queries drive the design. This is a complex chapter with difficult to teach subject matter. You could write an entire book on this one chapter. I do feel the author made a good effort here.Chapter 5 (Importing Data into Neo4j) - Put simply, this chapter describes multiple approaches on how to get data into a Neo4j database. Several tools are illustrated along with some pros and cons. This chapter focuses on tools and techniques and the author mentions the contents of this chapter are not an exhaustive list. However, there is plenty here to get you started.Chapter 6 (Use Case Example - Recommendations) - If you have read this far, you will find the author has done an excellent job making you very eager to read this chapter and the one to follow. In this chapter, you dive into the most familiar use case when the subject of graph databases comes up - recommender systems. The author sets the stage with a simple data model and demonstrates how Neo4j can issue queries against this data model to address recommendations regarding product purchases, brand loyalty, and social ties. The author provides informative Cypher queries along the way. Also included is what I regard as nice bonus material involving the same principles behind recommender systems - use case descriptions involving fraud detection, access control, and social networks. The author doesn't go into a lot of detail with these three use cases, but it was interesting to read about them.Chapter 7 (Use Case Example - Impact Analysis and Simulation) - I believe the goal of the author when writing this chapter was to expand the reader's horizons beyond the much more obvious and more commonly discussed use cases described in the prior chapter. In this chapter, the author begins by describing how Neo4j can help with business process management. In other words, we can model business processes as an interconnected system of workflows, software applications, departments, and physical assets such as buildings. Continuing along the same theme, the author demonstrates the usefulness of Neo4j as an impact simulation tool. One can better understand how an exogenous shock such as a price increase of a component can ripple through a complex product hierarchy and the associated product cost structure. Perhaps due to my own ignorance, I found this chapter to be the most esoteric part of the book, but I still found a lot of value from it. The moral of this story is that if a system can be mapped into a graph data structure, then Neo4j may be able to help.Chapter 8 (Visualizations for Neo4j) -From this point forward, the tempo of the book slows down. That's ok. For me, the intensity of the information from the prior chapters left me pretty full. In this chapter, the author goes through several tools, which includes JavaScript visualization libraries such as D3.js where you have to write code to solutions that have a graphical user interface. Even prior to reading this, I found that although graphical displays of large dense networks were cool to look at, it was really difficult to derive anything particularly actionable or insightful. The author even makes mention of that. This chapter provided me with some good information on visualization tools, but I still don't have a good understanding about how to use visualizations as a way to derive insights from complex networks. However, I don't hold that against the author.Chapter 9 (Other Tools Related to Neo4j) - In this chapter, the author surveys some data integration and business intelligence tools that can work with Neo4j databases. Of particular interest to me was learning about the existence of a JDBC driver for Neo4j. Finally, the chapter concludes with a brief discussion of graph modeling tools.Appendix A and Appendix B - The author provides some brief information on Neo4j community resources. What got my attention was the mention of the GraphGists website where people can post their own models and use cases. I plan on spending some time on that site. Also included is a helpful mention of a Cypher reference card along with a more detailed treatment of the Cypher key words. There isn't enough there to be a comprehensive explanation of Cypher, but there is enough to prepare you for a deeper dive.The author writes in a very clear style that is easy to read and follow. In fact, it was a pleasure to read. To be clear, this is an introductory book. You will come away with a firm foundation on graph databases and Neo4j as an implementation. I can say I would very much appreciate an advanced-level Neo4j book from the same author. If he decides to write such a book, I will certainly buy it.Finally, if the author's goal was to teach the basic concepts about graph databases via Neo4j in manner that peaks the reader's curiosity to learn more, then he was successful. If you already have some expertise with Neo4j, then this book may be much less of a benefit to you than it was for me. I maintain you should probably still check it out. I bet you will still learn some neat things you didn't know. This was a very well written book. I really enjoyed it and I recommend it. This book earns 5 stars from me.

One of the fascinated technical books I have ever read, while not difficult to understand (hint: Neo4J is well put together), it is very informative, yet elegantly written that to an able to learn on its own reader can open a whole slew of possibilities and opportunities.I need to step back and tell that I approached this book initially with some scepticism, and while I was on the very first pages of it one of my co-workers pryingly glanced over my shoulder and asked, what are you reading? I replied: “it is a book on a Graph database called Neo4J”. His reaction was, “I did not even know there are such a database type!”. As a database professional though I felt obliged to make myself aware and somewhat educated on alternative or niche databases, this is what I thought then, but now… read on.Ric Van Bruggen is definitely in love with Neo4J, but he is also a talented teacher, this book could much less of a technical reference than it is without him. Ric tastefully expands from the history and basics of Graph databases up to very practical, yet advance usages and ends the book with the tools ecosystem and even a neat Cypher ref-card. Besides, I think Cypher is the cornerstone to the successful operation of Neo4J. For the impatient reader, Cypher to Neo4J is as SQL to RDBMS’.The book reads like a story, it gradually builds your level of knowledge and at the same time makes you “feel at home” with Neo4J, it is a very welcoming database.My interest grew with every page I flipped.The database gets easily installed on any platform, and for the record, its graphing (visual) abilities are stunning! It can also output data in the tabular form. Yet, based on what Ric stated it is VERY fast – he processed 2.1 mil nodes in under 50 sec on a laptop (disclaimer: I did not run any Cypher code, but I am planning to).This book, especially its chapters 6 and 7 (use case examples), were of immense interest to me, the Recommendation Engine, Impact Analysis I loved both a lot. I think my case using Neo4J will be on finding the best driving path (based on more than 2 factors) that I want to experiment with.I figured, Neo4J is supported by a vast number of enterprise grade tools comprising a serious ecosystem in which Neo4J strives to take on new heights.After reading this book I must say the Graph Databases are seriously overlooked.Well done, Ric, Packt and etc. A 5 out of 5 from me!Take Neo4J for spin!

This book is a very good addition to the growing base of Neo4j documentation. Data import and use case examples are well written. Recommended.

Good book

This is a good introductory book for Neo4j for anyone considering a switch from RDBMS's. It gives a brief but very interesting history of Graph Theory and Databases, and how Graph Theory can solve some short comings of RDBMS's. The author presents solid use cases where a Graph Database could be preferred instead of a RDBMS, and vice versa. There is also some basic information about installation and setting up your environment, and basic Cypher usage. For advanced Cypher usage, you may want to look somewhere else.One of the short comings of the book is that there is no word about Operational perspective of Neo4j, scaling and maintenance, etc. Which are pretty important concerns when switching from an RDBMS. While I can understand that this is probably out of the scope of this introductory book, I would still expect at least some insight about scaling.Buy this book if you want to learn if a Graph Database is good for you or not, or if you need some help ideas about basic usage.