Tracking Latent Domain Knowledge - Mapping Scientific Frontiers: The Quest for Knowledge Visualization - page 239

Fig. 7.10 A landscape view of the Pathfinder case. Applications of Pathfinder networks are found
in a broader context of knowledge management technologies, such as knowledge acquisition,
knowledge discovery, and artificial intelligence. A majority of Pathfinder network users are
cognitive psychologists
of classic artificial intelligence articles, but it maintains a connecting point with
the global context through the “exit” landmark, which is the highest citation bar
half way down in the branch pointing to the lower right corner. This detailed local
structure shows more articles related to the use of Pathfinder.
Similarly, Table 7.4 shows leading articles in this latent knowledge structure. The
classification is more detailed than the one in the global context.
Figure 7.12 shows an extended branch from the main Pathfinder network. This
branch represents a new area of applying Pathfinder networks. In fact, this is the area
in which Pathfinder networks have been adapted for citation-based visualizations.
Tab le 7.5 reveals the fact that articles in this branch all have negative loadings
on factor one and are virtually absent from the remaining factors. This is interesting
because on the one hand, the first specialty provides a dimension that can account for
both the traditional applications of Pathfinders and the new branch of applications;
on the other hand, since documents in the new branch are so consistently classified
by factor loading, they can be treated as a sub-specialty.
Figure 7.13 shows a simple research function which lights up all the articles by
Schvaneveldt, a central figure in the development of Pathfinder network scaling. The
position of each lit article and the direction of the hosting branch provide insightful
information into the nature of the article and the branch.