20 of 20 people found the following review helpful
on October 10, 2013
The second edition of Mining the Social Web is not just an update of the previous edition (including Google+, GitHub, and Twitter API 1.1) but a new book. The book has been rethought in its entirety with a focus on pedagogy and practical use of the code. With the help of a virtual machine and IPython notebook (both made available by the author) it is possible to run the code without difficulty. The book includes a Twitter Cookbook section which is very useful if you want to mine Twitter. In my opinion this book is the best introduction to real-world programming in Python. It introduces many concepts and tools related to modern web-programming and data-mining. Additionally it gives you the tools and the code for querying social media APIs and analyzing your data in a meaningful way. Matthew Russell has realized a tour de force with the new edition of this book: introducing advanced programming concepts and tools in a pedagogic, accessible and practical way.
10 of 10 people found the following review helpful
on November 5, 2013
This book is extremely practical and has great code samples. It's easy to follow and fun! If you're interested in mining Twitter data, there is an (large) chapter focused entirely on reproducible code snippets that use the Twitter API.
11 of 12 people found the following review helpful
on January 22, 2014
Book review - Mining the Social Web, 2nd Edition by Matthew A. Russell, O'Reilly Media
Last year I read an article in Nature about Paul Erdős’s on the occasion of his 100th birthday. Outside mathematical circles Erdős is most known for the so called Erdős number. There are several different definitions of the Erdős number but according to Wikipedia it is defines as the "'collaborative distance' between a person and mathematician Paul Erdős, as measured by authorship of mathematical papers". So if you co-authored a paper with Erdős your Erdős number is 1. Your number will be 2 if you co-authored a paper with an author who wrote a paper directly with Erdős and so forth. Analyzing Erdős numbers is an application of social network theory and ever since I read the article I wanted to learn more about data mining applied to modern social media platforms. When researching for a book on this topic I came across Mining the Social Web and the books very practical approach convinced me to that this was the book I wanted to read.
Virtual Machine experience
The book is accompanied with a Virtual Machine experience that sets new standards for interactions between technical programming books and the code samples provided by the book. In no time you are up and running with the code samples in a IPython notebook that also can be edited and used as basis for your own data mining experiments. I would really love to see this approach adopted by other programming books.
The reader is gently guided through a software setup of VirtualBox and Vagrant and once these two programs have been installed it is just a matter of writing "vagrant up" in a terminal window and all of the necessary software used throughout the book will be installed and running in a virtual machine accessible through a web browser. Setting up the virtual machine might sound complicated but it is really quite easy. I tested the procedure for on both Mac and Windows and had no troubles getting the environment up and running in less than half an hour. And the really cool thing is that you don't have to install and manage a lot of dependencies yourself as well as you can delete everything afterwards just by deleting the virtual machine. The whole setup process is both described in the book and on videos found on the book's Github pages.
Some knowledge and experience with Python is required fully understand the code samples. If you have experience from other modern programming languages you should not have troubles understanding basic Python code. So the choice of Python cannot be considered as a barrier for reading the book.
The books does not cover social network theory in general nor graph theory so if you are looking for a book with a theoretical approach then this book is not for you. However most chapters in the book ends with a list of additional resources that can be used for further research.
This book is the best computer book I have read in several years. Social networks and data mining is a hot topic and reading Mining the Social Web will not only provide you knowledge about data mining but also supply practical code examples. In addition the books is an easy read and quite funny!
I review for the O`Reilly Reader Review Program and I want to be transparent about my reviews so you should know that I received a free copy of this ebook in exchange of my review.
11 of 12 people found the following review helpful
on November 8, 2013
Format: PaperbackVerified Purchase
I have purchased just about every book available on social media data mining/ analytics, including the first edition of this book. What Matthew Russell has done with this second edition is amazing. With the purchase of this book, you get a fully functional virtual machine (available via download on GitHub.) As updates are made to the code for the book, you can easily pull them from GitHub. This eliminates the countless hours you spend downloading, configuring, troubleshooting, wondering if you got the right version of the needed software, etc. Within minutes you can read the book and type the code samples. Actually, the code is already there, you simply enter in some key values and watch the code run. You can then morph the code and see the effects of your changes.
Mining the Social Web is exceptionally well written covering all major social media platforms. Mr. Russell is also very approachable and answers questions very quickly.
I really can't say enough good things about this book and how it sets the bar high for future technical books!
7 of 7 people found the following review helpful
on November 3, 2013
Great guidebook to acquiring and analyzing data from leading social media sites, including Twittter, Facebook, Google +, LinkedIn and GitHub along with other web tips and tricks. The iPython notebook approach provides turn key like method to run examples and check results in line, which accelerates and reinforces the topics.
Whether you are new to social media API's and want a straightforward way to ramp up learning and discovery of social mining techniques or more seasoned user, this book has it covered. Chapter formats and exercises make it easy to work a variety of topics and are laid out in easy to follow and execute fashion.
Highly recommend, so get the book and get started!
7 of 8 people found the following review helpful
on October 12, 2013
Mining the Social Web v2 is remarkable in terms of its simplicity as well as its depth. The author has focused on reducing friction to learning and executing traditionally difficult topics such as text mining and natural language processing. I already own the first version of MtSW, and between the new topics (LinkedIn, GitHub, Google+) and the new infrastructure (IPython, VirtualBox, etc) this is like a whole new book full of inspiration and ideas. The fact that a lot of this book is a significantly different than the first edition isn't surprising since the topic of the social web is evolving so rapidly.
The reason this is such an important book is that it teaches non-experts to build simple systems for making decisions on data that is constantly up-to-date. It's an end-to-end manual for continuously gathering data (e.g. Twitter API), analyzing data (e.g. Natural Language Processing), and presenting information (e.g. D3). By significantly reducing the barrier to building these systems, Matthew has increased the number of people on the planet that can provide data for making proper decisions . . . and data always beats opinions.
This is one of the rare books that does a great job of introducing deep technical topics AND providing an easy, practical implementation. Unlike a lot of tech books, MtSW makes it trivial to get started through a combination of Vagrant, VirtualBox, IPython Notebook, and GitHub such that you can have all the updated examples up and running within minutes. I'm much more of a practitioner (read: Hacker) than a computer scientist so this is exactly the right amount of technical detail to try out an idea. As an example of technical depth, the coverage of the Twitter API is exactly the proper amount of detail to understand how to pull out tweets and start using the data right away, without slogging through the parts of the API that you'll never need. Better yet, the examples in the book are implemented in IPython, so you can start using it right away and tweaking the code so you can learn it interactively.
4 of 4 people found the following review helpful
on October 10, 2013
If you want to start crawling data from the web this is THE BOOK. They have a lot of pratical exelples in python and open your mind to this chalenge area. The second edition now includes other social medias (G+ etc) with exemples (for me the examples are the BESTE part of the book). You don't need to read the full book (just the social media that you are interested), but i highly recomend! I could't finish my PhD with out it!
9 of 11 people found the following review helpful
on November 3, 2013
Format: Kindle Edition
Why in the world would anyone want to mine data from social websites, you may be asking yourself just about now. Good question. Suppose you were in the process of creating a product, but at the same time you are curious as to which niche it would fit into. You may also be curious as to which niche is the most financially beneficial for your product, as well as perhaps you should tweak it to maximize your particular niche after mining the web for this data.
Who would benefit from this product the most? And best of all, which social websites do your prospective buyers frequent the most. Is it Facebook? What about Twitter? Do they have a membership on LinkedIn? Are they a member of Google+? Regardless of where they may be, there is a good chance that your data mining will pay off.
There is plenty of example code, which makes use of the Python language. There is also IPython Notebook which is an interactive Python interpreter which gives you a notebook like experience from your web browser. With a few clicks from within IPython Notebook, you can be well on your way to learning more about the users of social websites than you might have ever thought possible.
A part of the paragraph on IPython is paraphrased from the books itself. I would definitely recommend this book to others. It looks great on my Kindle Fire HD.
3 of 3 people found the following review helpful
on July 30, 2014
I was initially skeptical that a book with a relatively small number of reviews written by an author I had never heard of could have such a high number of 5-star reviews -- all of which were rated as 'helpful' by nearly everyone who rated the reviews. After expressing my doubts about the veracity of the reviews, the author kindly shipped me two signed copies of his book: one to keep and one to give to a friend. Now that I have read it, I understand why so many people give it 5 stars.
I'm one of those people who reads programming books to learn what is possible and then hires a code monkey to implement it. Programmers are a cheap and abundant resource, but Mining the Social Web makes you want to jump in and build something yourself. As other reviews have mentioned, a base level of prior programming experience is necessary to get the most out of this book. However, I spent just two days learning basic Python syntax and logic while I was waiting for the book to arrive, and found that my knowledge was sufficient to follow the examples in the book.
Most of the book's immediate value comes from walking through practical examples of whatever you need to implement at the moment. If, for instance, you want to collect and analyze Twitter data, the first chapter shows you how to use Twitter's API and walks through specific examples. The book covers a wide range of topics; just glance at the chapter titles in the table of contents to see exactly what it covers.
The book also contains general knowledge about data mining that has broad application across platforms. For example, it gives detailed explanations for various algorithms, such as k-means clustering, that can be used in a variety of settings. These algorithms are usually buried within chapters on a specific social media platform, so it makes it worthwhile to read the chapter on, say, LinkedIn, even if you don't intend to mine LinkedIn data.
I recommend Mining the Social Web (2nd edition) to anyone who builds web applications or just wants to do some cool stuff with online data. I encourage prospective buyers to read the author's blog post (http://miningthesocialweb.com/2013/08/24/reflections-on-authoring-a-minimum-viable-book/) about the making of the second edition -- it shows how responsive Russell was to comments on the first edition and why the second edition is flooded with 5-star reviews.
3 of 3 people found the following review helpful
on January 13, 2014
"Mining the Social Web, 2e" is a terrifically insightful, practical book that every serious data enthusiast should read. The book represents almost rewrite from the 1st edition.
What sets this book apart from other technical books is the very obvious care and time the author has spent creating it. In order to get the interesting algorithmic work you always have to interact with myriad APIs and libraries, and this is usually a complex, painful process that works out to be harder than the issue you are attempting to understand. Matthew has abstracted that whole layer of trouble by providing all of the examples and code via a virtual machine serving IPython notebooks. This means that there are no libraries or version conflicts with which to contend, so your focus stays on the interesting subject matter of the book. Because of this abstraction Matthew is able to explore a myriad of different real world systems from Twitter to Facebook to email, all the time allowing the reader to succeed at the analysis by having the plumbing in place already.
The examples are in depth without veering in mathematical arcana, the focus is on what is possible now, and how to do it today. The recurring themes are social graphs, natural language processing and machine learning along with some visualisation. Each section of the book uses these tools on different real data sets showing their correct application and interpretation. Complex topics such as NLP as built up over several sections with clear and generous explanation.
I do not envisage attempting a problem in this area without first re-reading the relevant sections from this book.