What do you think?
Rate this book
Graph-Powered Machine Learning
Graph-Powered Machine Learning introduces you to graph technology concepts, highlighting the role of graphs in machine learning and big data platforms. You’ll get an in-depth look at techniques including data source modeling, algorithm design, link analysis, classification, and clustering. As you master the core concepts, you’ll explore three end-to-end projects that illustrate architectures, best design practices, optimization approaches, and common pitfalls. Author Alessandro Negro’s extensive experience building graph-based machine learning systems shines through in every chapter, as you learn from examples and concrete scenarios based on his own work with real clients!
- GenresArtificial Intelligence
475 pages, Paperback
Published January 1, 2021
Ratings & Reviews
Friends & Following
Create a free account to discover what your friends think of this book!
Community Reviews
Displaying 1 - 4 of 4 reviews
September 14, 2020
Frustrating book that is very repetitive and yet tantalising - you constantly feel like you’re on the verge on enlightenment only to be forestalled page after page. Long passages explaining the motivating scenarios seem to be reworded in each chapter. Examples are in Python, scikit-learn and Neo4j.
Caveat: This book is still in draft form and has been made available as part of Manning’s Early Access Program (MEAP). In particular this means that there has been no copy editing or seeming any proof reading.
Caveat: This book is still in draft form and has been made available as part of Manning’s Early Access Program (MEAP). In particular this means that there has been no copy editing or seeming any proof reading.
August 19, 2026
After reading Graph Neural Networks in Action, this book was a frustrating experience. There are endless walls of text, images are repeated multiple times, and the overall structure often hides the important ideas rather than revealing them.
There is certainly useful material here. In particular, it is interesting to see how data can be transformed into graphs and how a graph database can do much of the heavy lifting. Unfortunately, these insights are buried beneath constant repetition and excessive explanation. Even the introduction to a new part spends far too much time recapping what came before. Do we really need a full-page reminder of what happened in Part 2 at the beginning of Part 3? I don't think so.
Even more frustrating is the author's tendency to over-explain the most basic details. A particularly striking example is a one-sentence quote from Van Vlasselaer defining fraud, which is followed by roughly 600 words dissecting the meaning of its individual verbs. If the intention is to bridge the gap between the legal definition of fraud and the challenges it creates for machine learning, a few carefully chosen sentences could make that point much more clearly. While this example is extreme, it unfortunately captures one of the book's biggest problems: important ideas regularly disappear beneath unnecessary wordiness.
The treatment of NLP suffers from much the same issue. It is one of the most drawn-out and ultimately least useful explanations of NLP I have encountered. There may well be benefits to the graph-based approach the author presents, but they are difficult to identify amid the lengthy explanations. The fact that chapters repeatedly need a separate "Advantages of the graph approach" section is almost telling in itself: if those advantages were demonstrated clearly throughout the chapter, they would not need to be spelled out again afterward.
There are worthwhile ideas in this book, particularly around representing data as graphs and taking advantage of graph databases. Unfortunately, they are buried under repetition, excessive detail, and a structure that makes the material harder to understand than it needs to be.
Overall, this was one of the most disappointing Manning books I have read.
There is certainly useful material here. In particular, it is interesting to see how data can be transformed into graphs and how a graph database can do much of the heavy lifting. Unfortunately, these insights are buried beneath constant repetition and excessive explanation. Even the introduction to a new part spends far too much time recapping what came before. Do we really need a full-page reminder of what happened in Part 2 at the beginning of Part 3? I don't think so.
Even more frustrating is the author's tendency to over-explain the most basic details. A particularly striking example is a one-sentence quote from Van Vlasselaer defining fraud, which is followed by roughly 600 words dissecting the meaning of its individual verbs. If the intention is to bridge the gap between the legal definition of fraud and the challenges it creates for machine learning, a few carefully chosen sentences could make that point much more clearly. While this example is extreme, it unfortunately captures one of the book's biggest problems: important ideas regularly disappear beneath unnecessary wordiness.
The treatment of NLP suffers from much the same issue. It is one of the most drawn-out and ultimately least useful explanations of NLP I have encountered. There may well be benefits to the graph-based approach the author presents, but they are difficult to identify amid the lengthy explanations. The fact that chapters repeatedly need a separate "Advantages of the graph approach" section is almost telling in itself: if those advantages were demonstrated clearly throughout the chapter, they would not need to be spelled out again afterward.
There are worthwhile ideas in this book, particularly around representing data as graphs and taking advantage of graph databases. Unfortunately, they are buried under repetition, excessive detail, and a structure that makes the material harder to understand than it needs to be.
Overall, this was one of the most disappointing Manning books I have read.
November 28, 2020
I read prerelease version of the book...
good overview of the application of the graph approach to solving the problems traditionally solved with machine learning - recommendations, etc.
The biggest drawback from my side - sometimes it too wordy - you need to read a lot (sometimes is not completely related to the topic) before getting to the solving of the real problems.
good overview of the application of the graph approach to solving the problems traditionally solved with machine learning - recommendations, etc.
The biggest drawback from my side - sometimes it too wordy - you need to read a lot (sometimes is not completely related to the topic) before getting to the solving of the real problems.
April 18, 2025
Didn't get to the good stuff and repeated information over and over again.
Displaying 1 - 4 of 4 reviews