Network Science

by Albert-László Barabási, Márton Pósfai (Primary Contributor)

Networks are everywhere, from the internet, to social networks, and the genetic networks that determine our biological existence. Illustrated throughout in full colour, this pioneering textbook, spanning a wide range of topics from physics to computer science, engineering, economics and the social sciences, introduces network science to an interdisciplinary audience. From the origins of the six degrees of separation to explaining why networks are robust to random failures, the author explores how viruses like Ebola and H1N1 spread, and why it is that our friends have more friends than we do. Using numerous real-world examples, this innovatively designed text includes clear delineation between undergraduate and graduate level material. The mathematical formulas and derivations are included within Advanced Topics sections, enabling use at a range of levels. Extensive online resources, including films and software for network analysis, make this a multifaceted companion for anyone with an interest in network science.

Genres: Science, Computer Science, Mathematics, Textbooks, Nonfiction, Sociology, Academic

475 pages, Hardcover

First published January 1, 2016

About the author

Albert-László Barabási is a physicist, best known for his work in the research of network science. A Hungarian born native of Transylvania, he received his Masters in Theoretical Physics at the Eotvos University in Budapest, Hungary and was awarded a Ph.D. three years later at Boston University. Barabási is the author of six books, including the forthcoming book "The Formula: The Science of Success." His work lead to the discovery of scale-free networks in 1999, and proposed the Barabási-Albert model to explain their widespread emergence in natural, technological and social systems, from the cellular telephone to the WWW or online communities.

Barabási is both the Robert Gray Dodge Professor of Network Science and a Distinguished University Professor at Northeastern University, where he directs the Center for Complex Network Research, and holds appointments in the Departments of Physics and Computer Science, as well as in the Department of Medicine, Harvard Medical School and Brigham and Women Hospital, and is a member of the Center for Cancer Systems Biology at Dana Farber Cancer Institute.

Barabási is a Fellow of the American Physical Society. In 2005 he was awarded the FEBS Anniversary Prize for Systems Biology and in 2006 the John von Neumann Medal by the John von Neumann Computer Society from Hungary, for outstanding achievements in computer-related science and technology. In 2004 he was elected into the Hungarian Academy of Sciences and in 2007 into the Academia Europaea. He received the C&C Prize from the NEC C&C Foundation in 2008. In 2009 APS chose him Outstanding Referee and the US National Academies of Sciences awarded him the 2009 Cozzarelli Prize. In 2011 Barab‡si was awarded the Lagrange Prize-CRT Foundation for his contributions to complex systems, awarded Doctor Honoris Causa from Universidad PolitŽcnica de Madrid, became an elected Fellow in AAAS (Physics) and is an 2013 Fellow of the Massachusetts Academy of Sciences.

Reviews

[Gregory]

This book is freely available at http://barabasi.com/networksciencebook

[Andreia Pereira · Rating 4 out of 5]

Such an easy read. One of the reasons I loved my Network Science course so much this semester.

Really nice explanation of computational epidemiology too.

[Fathin]

Excellent as an introductory reading for understanding network analysis. Great that it is available in Open Access with teaching slides and raw data!

[Riccardo · Rating 5 out of 5]

A thorough introduction to network analysis with a strong and refreshing focus on the mechanisms at play, beyond descriptive approaches. As a cognitive scientist I guess I’ll always have a physics envy, an awe of the elegant mathematical approach carving phenomena at minimal joints. The book is wonderful in guiding you through this approach and the thinking and math involved. As a downside, I would have liked a more problem-based approach, with worked out examples of how to apply the math and approach to real problems (tho a valiant attempt is made in the homework).

[Mik Zlatin · Rating 5 out of 5]

I need more of this type of book. It was a perfect blend of the type of math I love with applications and the relevance of different models. 10/10

No but seriously if anyone has similar book recommendations to this, please let me know!

[Jacob Williams · Rating 4 out of 5]

This was one of the textbooks for the Network Science course I just took at Georgia Tech. I mostly don’t care about the math proofs, but the high-level concepts are interesting. The idea is that when something can be modeled as a set of nodes linked by edges—examples studied in the book include the Internet, the network of citations among academic papers, the graph of who calls who on cell phones, and the map of binding interactions among proteins—we can learn many useful things about it just by studying the the structure of that graph, ignoring what the nodes and edges actually represent.

Barabási’s key claim is that in most real-world networks, the node degrees follow a power-law distribution. That is, if you count how many connections each node has (its “degree”), lower-degree nodes tend to be much more common than higher-degree nodes, but you can still expect to find some nodes with arbitrarily high degrees. This is not what you’d expect to find in a purely random network; for random networks, the degree distribution is binomial, so most nodes should have similar numbers of connections and extreme outliers are unlikely. Networks with a power-law degree distribution are also called “scale-free” networks.

The degree distribution has fascinating implications for how vulnerable a network is to attacks and random failures. How many nodes have to fail before some portion of the network becomes incapable of reaching some other portion?

- If the failures are happening randomly, they cause significantly more damage to a random (binomial degree distribution) network than to a scale-free (power-law degree distribution) network. Scale-free networks tend to contain some “hub” nodes that are extremely helpful in keeping the network connected, so unless you have the extreme bad luck of losing those hub nodes, the loss of a few random nodes doesn’t make much difference.
- But for the same reason, scale-free networks are very vulnerable to intelligent attacks, even moreso than random networks are. If an attacker knows which nodes are hubs and targets those first, the network can fall apart rapidly.

This book opens with a somewhat unusual “personal introduction” chapter that I appreciated—it describes some of the risks the author and his collaborators took, and disappointments he endured, in the long journey toward getting the academic community to see the value in studying this topic.

(crosspost)

[David Goh · Rating 4 out of 5]

Definitely catches you up to speed on a lot of network stuff with a main focus on scale free / preferential attachment mechanisms. Which to be fair is interesting- who would have guessed that Zipf's Law/the Pareto Principle shows up in networks too? But the scope is certainly limited: algebraic tools like spectral graph theory, which is my personal favorite, are not covered. And who would have guessed that the field would have moved so fast in the ensuing decade, where many contemporary researchers dispute / call for more rigor in testing the universality of the scale free property, Barabási's dear own discovery?

Read most of this for research, and I'm listing it here to make it through my Goodreads challenge I guess.

[Dezső Módos · Rating 5 out of 5]

If you are interested in Network science you must read this book. It discuss network topology and basic network metrics and network types such as Erdős Rényi random network or Barabási Albert graphs. Note it is not a mathematical book so you do not need deep maths to understand it, but also it does not have the maths book thesis proof system. It is a much more empirical book.

[Mostafa Alkady · Rating 5 out of 5]

The perfect book to start learning the basics of network science and its applications. Easy to follow and available online for free. For deep dives consult Networks by Mark Newman.

[Nadvornix · Rating 4 out of 5]

I recommend for people to read something about network science, since it is an important theory for understanding the world. This one is fine with focus on applications (you want network science for its applications, not for its theory), even through a bit ego-centric.

[Michiel · Rating 4 out of 5]

Interesting reference for network science. Though focussed on the classical stuff, this is pleasant to read with many real-world examples.

[Ni Khan · Rating 5 out of 5]

A reference book for social network analysis.

[Qurban · Rating 5 out of 5]

Riyaziyyatdan psixologiyaya bir çox sahəni özündə birləşdirən mövzu olan sosial şəbəkə haqda fundamental dərslikdir.

Müəllifin saytında kitabın bir neçə dildə elektron versiyası var.

[Matias Andrade · Rating 4 out of 5]

great book, solid intro. Recommend reading Statistical Mechanics journal 50 page summary on research to catch up to modern results.