The Dynamics of Scientific Knowledge - Mapping Scientific Frontiers: The Quest for Knowledge Visualization - page 38

that provide the base map of a thematic map. Indeed, thematic maps provide a
prosperous metaphor for a class of information visualization known as information
landscape. Notable examples include ThemeView (Wise et al. 1995 ) and Bead
(Chalmers 1992 ).
ManyEyes is a more recent example. It is a 'social kind of data analysis' in the
words of its designers at the formerly IBM's Visual Communication Laboratory.
ManyEyes enables many people to have a taste of what is like to create your
own information visualization that they would otherwise have no such chance
at all. The public-oriented design significantly simplifies the entire process of
information visualization. Furthermore, ManyEyes is indeed a community-building
environment in which one can view visualizations made by other users, make
comments, and make your own visualizations. These reasons alone would be
enough to earn ManyEyes a unique position in the development of information
visualization. ManyEyes and Wikipedia share some interesting characteristics—
both tap in social construction and both demonstrate emergent properties of a
self-organizing underlying system.
Modeling and visualizing intellectual structures from scientific literature have
reached a new level in terms of the number of computer applications available,
the number of researchers actively engaged in relevant areas, and the number of
relevant publications. Traditionally, the scientific discipline that has been actively
addressing issues concerning science mapping and intellectual structure mapping
is information science. Information science itself constitutes of two sub-fields:
information retrieval and citation analysis. Both information retrieval and citation
analysis take the widely accessible scientific literature as their input. However,
information retrieval and citation analysis concentrate on disjoint sections of a
document. Information retrieval focuses on the bibliographic record of a document,
such as title and keyword list, and/or the full-text of a document, whereas citation
analysis focuses on referential links embedded in the document, or those appended
at the end of the document. The ultimate challenge for information visualization is to
invent and adapt powerful visual-spatial metaphors that can convey the underlying
semantics.
Information retrieval has brought many fundamental inspirations and challenges
to the field of information visualization. Our quest aims to demonstrate that
science mapping goes beyond information retrieval, information visualization, and
scientometrics. It becomes a unique field of study on its own and yet it has the
potential to be applicable to a wide range of scientific domains. Our focus is on the
growth of scientific knowledge and what are the key problems to solve and what
are the central tasks to support. Instead of focusing on locating specific items in
scientific literature, we turn to higher levels of granularity - scientific paradigms
and their movements in scientific frontiers.
Visual analytics can be seen as the second generation of information visualiza-
tion. It has transformed not only how we visualize complex and dynamic phenomena
in the new information age, but also how we may optimize analytical reasoning and
make sound decisions with incomplete and uncertain information (Keim et al. 2008 ).
Today's widespread recognition of the indispensable value of visual analytics as a