The Dynamics of Scientific Knowledge - Mapping Scientific Frontiers: The Quest for Knowledge Visualization - page 3

maturing field of visual analytics provides a promising direction of pursuit. Visual
analytics can be seen as the second generation of information visualization. The
first-generation information visualization aims to gain insights from data that
may not readily lend itself to an intuitive visual and spatial representation. The
second-generation visual analytics makes it explicit that the goal is to support
evidence-based reasoning and decision making activities.
Due to its very nature, science mapping involves a broad variety of units
of analysis at different levels of granularity. The notions of macro-, meso-, and
microscopic levels can be helpful to clarify these units, although the precise
definition of these levels themselves is subject to a debate. For example, at a
macroscopic level, we are interested in the structure and dynamics of a discipline
and the entirety of a scientific community; we may even want to study how multiple
disciplines interact, for example, in the study of interdisciplinarity. Importantly,
many studies have suggested that interdisciplinary activities may play an essential
role in the development of science. Boundary-spanning activities in general may
indeed hold the key to scientific creativity.
At a lower level of aggregation, the meso scale often refers to a system of groups.
In other words, the unit of analysis at this level is groups. The existence of this
level implies that the macro and societal level is heterogeneously distributed. It is
not evenly distributed. Science mapping at this level corresponds to the study of
paradigms, including a thematic thread of research that could rise and fall over
time. At the even lower microscopic level, the units of analysis include individual
scientists and particular approaches to specific topics and solutions to specific
problems.
Scientific literature provides a wide range of options for researchers to choose
their units of analysis. For example, subject categories in the Web of Science have
been used to represent sub-disciplines. Cited references have been used to indicate
concept symbols. Word occurrence patterns have been used to model underlying
topics. What is special about the new generation of science mapping efforts is the
realization of the profound role of handling large-scale and heterogeneous sources
of streams of data so that one can explore the complexity of the development of
scientific knowledge from a broad range of perspectives. This realization in turn
highlights the significance of studying how distinct perspectives interact with each
other and how we may improve our understanding of science in terms of hindsight,
insights, and foresights.
The remarkable progress in science mapping in the recent few years is one of the
series revivals of what was pioneered in the 1960s and 1970s. The most seminal
works in information science include the contribution of Derek de Solla Price
(1922-1983), namely his Networks of Scientific Papers (Price 1965 ), Little Science,
Big Science (Price 1963 ), and Science since Babylon (1961). In Little Science, Big
Science , Price raised a series of questions that subsequently inspired generations
of researchers in what is now known as the science of science : Why should we
not turn the tools of science on science itself? Why not measure and generalize,
make hypotheses, and derive conclusions? He used the metaphor of studying the
behavior of gas in thermodynamics to illustrate how the science of science could