Visual Analytics - Mapping Scientific Frontiers: The Quest for Knowledge Visualization - page 317

Fig. 9.1 A screenshot of GeoTime (Reprinted from Eccles et al. 2008 )

of scientific paradigms in the sense defined by Thomas Kuhn. The primary source
of input for CiteSpace is a body of scientific literature, namely bibliographic records
from the Web of Science or full-text versions of publications.
The general assumption is that the study of such input data will allow us to
address two fundamental questions that systematic reviews and surveys would intent
to address:

1. What is the persistent core of the literature?
2. What are the transient trends that have appeared or are emerging in the literature?

The persistent core of a body of literature corresponds to the intellectual base of
a field of study. The transient trends correspond to scientific frontiers. On the other
hand, researchers have realized that scientific knowledge can be seen as the constant
movement of scientific frontiers. The state of art today may or may not survive the
future. Only the time can tell whether an exciting new theory will have its position
in the history of science.
We use co-citations of references as the basic organizing mechanism. In other
words, we construct a global structure from local details. Each individual scientist
or domain expert provides their input as they publish their work in the literature. As
they cite previously published works in the literature, they leave their footprints that
carry information about their preferences, intents, criticisms, and interpretations. In
this way, citations provide a valuable source of information for us to identify and
measure the value of a scientific idea, a discovery, or a theory.