Crossmaps: visualization of overlapping relationships in collections of journal papers.

[Steven Morris]

Hello folks,

When mapping entities such as authors, papers, or references in a
collection of journal papers there is usually a considerable amount of
overlap in relations among the clusters of entities.  This can lead to
severe problems when trying to interpret MDS maps of such entities.
Rather than fighting or ignoring the problem of interpreting overlapping
relations it is possible to take a "stop worrying and learn to love the
bomb" approach to the problem.  Crossmaps are a simple technique that
maps relations among groups of entities (as opposed to the usual method
of using MDS to map the entities themselves), and allows visualization
and interpretation of overlapping relations.

The technique is explained in a PNAS paper published this week:

Morris, S. & Yen, G., (2004) "Crossmaps, visualization of overlapping
relationships in collections for journal papers." Proceedings of the
National Academy of Sciences of the United States of America.
<http://www.pnas.org/cgi/reprint/0307604100v1.pdf>

This paper reports on material presented at the NAS Sackler Colloquium
on "Mapping Knowledge Domains" held May 9-11, 2003 in Irvine,
California.  The original poster presentation for this materical can be
seen here
<http://samorris.ceat.okstate.edu/web/networks/sackler_poster.ppt>, and
constitutes a good summary of the information presented.  Readers
wishing to explore the example collection of papers on "complex
networks" presented in the paper can do so here
<http://samorris.ceat.okstate.edu/web/networks/>.  Links labeled  P, R,
AP, AR, JP and JR, on the left side of the timeline plot can be used to
get lists of papers, and ranked lists of references, paper authors,
reference authors, paper journals and reference journals in the various
clusters of papers in the collection.   A second example,  presented at
the colloquium but not included in the paper because of space
limitations demonstrates the use of crossmapping on a factor matrix from
an author co-citation analysis taken from:

H. White and K. W. McCain, "Visualizing a discipline: an author
co-citation analysis of information science, 1972-1995," Journal of the
American Society for Information Science and Technology, vol. 49, pp.
327-355, 1998.

The plot from that example can be found here
<http://samorris.ceat.okstate.edu/web/info_sci/wm.htm>.  A comparison
study of Information Science authors, using about 14,000 abstracts from
12 Information Science journals is  shown here
<http://samorris.ceat.okstate.edu/web/info_sci/default.htm>.

In my view, one of the most interesting aspects of the crossmaps paper
is the entity-relation diagram of paper collections shown in Figures 1
and 2 of the paper.   In the bibliometrics literature there are a
confusing number of entity-types such as papers, references, terms,
authors, and journals, and, usually with no distinction between
reference authors and paper authors, reference journals and paper
journals.   There are usually four entity occurence distributions often
discussed, Lotka's Law, Bradford's Law, Zipf's Law and the reference
power law.  A large number of co-occurence measures appear as well,
co-citation, bibliographic coupling, author co-citation, co-authorship
and co-words, for example.  Additionally, many types of entity clusters
are discussed: research fronts, co-citation clusters, author co-citation
clusters and collaboration groups.  The entity-relation diagram shown in
Figure 1 of the paper,  maps all of these entity-types, distributions,
co-occurrences, and entity clusters in a logical and simple way.  In my
opinion. :-)

Simplified versions of this diagram can be used as web  navigation aids
for showing results of bibliometrics studies, such as shown here
<http://samorris.ceat.okstate.edu/web/default.htm>, where a diagram
provides links to plots of entity  occurence distributions for a series
of paper collections.  (The links on the diagram are underlined blue
text.)  Note that journal co-citation is missing from the diagrams, but
can be added by drawing lines between papers and reference journals,
adding symmetry on the diagram with author co-citation.

Any feedback or suggestions about this work would be greatly appreciated.

Thanks,

S. Morris

--
---------------------------------------------------------------
Steven A. Morris                            samorri at okstate.edu
Electrical and Computer Engineering        office: 405-744-1662
202 Engineering So.
Oklahoma State University
Stillwater, Oklahoma 74078
http://samorris.ceat.okstate.edu

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.asis.org/pipermail/sigmetrics/attachments/20040206/6cc34d52/attachment.html>