Order this book
Adam G. Dunn
Centre for Health Informatics, Australian Institute of Health Innovation
Vladimir Batagelj is one of the original developers of Pajek (Batagelj, 1998), a well-known software tool for analysing and visualising networks that was created nearly twenty years ago. The purpose of the book at hand is to guide readers through the entire process of investigating questions that can be answered using information about networks that may be spatial or temporal. This means that the authors go well beyond just surveying methods, covering the development of a research question, the practicalities of cleaning data, and identifying the right methods to answer the question. This is the first of many points of departure that sets this book apart from others in the area.
I was concerned that the book might be one long advertisement for Pajek but this was not the case. The book remains readable and teachable even without running through the examples in Pajek in parallel. The second and third chapters cover basic and more advanced methods that are standard in network science, respectively. These chapters are especially dense with information – Dunbar’s number gets one sentence.
In the remaining chapters, Batagelj et al. teach the reader how to look at application domains in computational social science – both hypothesis-driven and explanatory studies. Chapters 4 to 9 cover bibliographic and patent citation networks, US Supreme Court citation networks (following Fowler and Jeon, 2008), football player movements over their careers, and the different ways to explain the spatial diversity of the United States.
It is refreshing to see the rigorous kind of computational social science that properly considers the context and prior work rather than jumping in to apply a favourite tool or method. The authors have done this consciously, concluding: “(U)sing methods without substance seems an empty endeavor when understanding network phenomena is the central concern.”
It is this questioning of the status quo that permeates the book. The authors cover controversial topics like the patriarchal, colonial, and sometimes racist nature of Supreme Court decisions as an application domain. The myths of player trading in football are challenged. Yet beyond this, the very nature of the book is to present network science in a way that does not neatly fit into either the sociology or physics tribes, instead embedding the methods into the basic mathematical structures that are common to both.
The decades of shared practical experience among the authors are apparent. For example, it is notoriously difficult to squeeze the visualisation of large networks into the margins of a book without making them look like ugly hairballs that take up too much room and fail to demonstrate any real aspect of the structure. If readers have attempted this task before, they will realise how much of an accomplishment it is that Batagelj and co-authors have managed to present large networks, with legible labels, and in ways that impart relevant information about the questions being asked.
I wanted to see more examples of spatial networks beyond what was presented in Chapter 9, which was really the first complete chapter focused on spatial analysis. Examples of where information flows through, and is modified by, spatial constraints would not look out of place in the book. For example, the flows of populations and cultures over time have been popular topics in JASSS and I would have appreciated seeing a chapter in this area.
In summary, Batagelj et al. have produced a clear and accessible textbook, balancing symbolic maths, code, and visual explanations. The authors’ enthusiasm for the subject matter makes it enjoyable to read despite the density of methods in the introductory chapters. Understanding Large Temporal Networks and Spatial Networks is appropriate for undergraduates and postgraduates with some background in algebra, set theory, and data structures. This is not another book surveying the most recent methods in community structure analysis or epidemic modelling over social networks. Network Science (Barabási, 2015) is a slicker and more specialised example of that. This book is for readers who are looking to learn more than just the methods but want to understand the entire investigative process of addressing spatial and temporal problems in network science.
BATAGELJ, V. and Mrvar, A. (1998). Pajek – program for large network analysis. Connections, 21(2), 47-57.
BAVELAS, A. (1948). A mathematical model of group structures. Human Organization, 7(3), 16-30.
FOWLER, J. H. and Jeon, S. (2008). The authority of Supreme Court precedent. Social Networks, 30(1), 16-30.
Return to Contents of this issue
© Copyright JASSS, 2015