Mapping Scientific Frontiers: The Quest for Knowledge Visualization
By Chaomei Chen.
Visualization of scientific frontiers is a relatively new field, yet it has a long history and many predecessors. The application of science to science itself has been undertaken for decades with notable early contributions by Derek Price, Thomas Kuhn, Diana Crane, Eugene Garfield, and many others. What is new is the field of information visualization and application of its techniques to help us understand the process of science in the making. In his new book, Chaomei Chen takes us on a journey through this history, touching on predecessors, and then leading us firmly into the new world of Mapping Scientific Frontiers. Building on the foundation of his earlier book, Information Visualization and Virtual Environments, Chen's new offering is much less a tutorial in how to do information visualization, and much more a conceptual exploration of why and how the visualization of science can change the way we do science, amplified by real examples.
Chen's stated intents for the book are:
On all these counts, he succeeds.
Chen's book can be broken into two parts which focus on the first two purposes stated above.
The first, consisting of the initial four chapters, covers history and predecessors. Kuhn's theory of normal science punctuated by periods of revolution, now commonly known as paradigm shifts, motivates the work. Relevant predecessors outside the traditional field of information science such as cartography (both terrestrial and celestial), mapping the mind, and principles of visual association and communication, are given ample coverage. Chen also describes enabling techniques known to information scientists, such as multi-dimensional scaling, advanced dimensional reduction, social network analysis, Pathfinder network scaling, and landscape visualizations. No algorithms are given here; rather, these techniques are described from the point of view of enabling 'visual thinking'. The Generalized Similarity Analysis (GSA) technique used by Chen in his recent published papers is also introduced here.
Information and computer science professionals would be wise not to skip through these early chapters. Although principles of gestalt psychology, cartography, thematic maps, and association techniques may be outside their technology comfort zone, or interest, these predecessors lay a groundwork for the "visual thinking" that is required to create effective visualizations. Indeed, the great challenge in information visualization is to transform the abstract and intangible into something visible, concrete, and meaningful to the user.
The second part of the book, covering the final three chapters, extends the mapping metaphor into the realm of scientific discovery through the structuring of literatures in a way that enables us to see scientific frontiers or paradigms. Case studies are used extensively to show the logical progression that has been made in recent years to get us to this point. Homage is paid to giants of the last 20 years including Michel Callon for co-word mapping, Henry Small for document co-citation analysis and specialty narratives (charting a path linking the different sciences), and Kate McCain for author co-citation analysis, whose work has led to the current state-of-the-art.
The last two chapters finally answer the question"What does a scientific paradigm look like?" The visual answer given is specific to the GSA technique used by Chen, but does satisfy the intent of the bookto introduce a way to visually identify scientific frontiers. A variety of case studies, mostly from Chen's previously published worksupermassive black holes, cross-domain applications of Pathfinder networks, mass extinction debates, impact of Don Swanson's work, and mad cow disease and vCJD in humanssucceed in explaining how visualization can be used to show the development of, competition between, and eventual acceptance (or replacement) of scientific paradigms. Although not addressed specifically, Chen's work nonetheless makes the persuasive argument that visual maps alone are not sufficient to explain "the making of science" to a non-expert in a particular field. Rather, expert knowledge is still required to interpret these maps and to explain the paradigms. This combination of visual maps and expert knowledge, used jointly to good effect in the book, becomes a potent means for explaining progress in science to the expert and non-expert alike.
Work to extend the GSA technique to explore latent domain knowledge (important work that falls below the citation thresholds typically used in GSA) is also explored here. This is extremely important in that this technique can allow new theories to be spotted and publicized within a community long before they become popular, and thus may accelerate debate and discovery in science.
Dr. Chen notes that continued progress in the ability to map science effectively will require the combined efforts of those from many disciplines including science philosophy, scientometrics, knowledge discovery, data mining, and visualization. To that list, I would add the domain experts in all branches of science that would benefit from insertion of knowledge beyond their own specialties. The work of Don Swanson on linking non-interactive literatures, which is given a brief overview in the final chapter, is a strong reminder that all scientists should be looking outside their specialties. It is my hope that Chen's final intent for the book, to stimulate interdisciplinary research, will indeed come to pass. Information in the book is well compartmentalized, with references at the end of each chapter, which will appeal to those wanting to update their knowledge with the most recent information.
 Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company, for the United States Department of Energy's National Nuclear Security Administration under contract DE-AC04-94AL85000.