Social, and other, Networks
Networks can be considered to be very large graphs with several hundreds
of nodes.
The can be undirected (symmetric), directed (asymmetric) or mixed.
While there are many statistical methods of analyzing networks, they
tend to yield only global properties.
Closed set analysis can reveal much finer detail.
One analytic technique is reduction.
All nodes are closed in an irreducible network.
The irreducible ``core'' of any network is unique (upto isomorphism).
In this irreducible core, every node is a part of a chordless circuit of
length 4, or more.
Thus they are the exact antithesis of chordal graphs.
They are
often an order of magnitude smaller.
Consequently, they can be used to identify local network features of interest.
-
[P11] J.L. Pfaltz,
Mathmetical Continuity in Dynamic Social Networks,
3rd International Conf. on Social Informatics, SOCINFO 2011 ,
October 2011,
LNCS #6984, 36-50.
-
[PT12a] J.L. Pfaltz
Finding the Mule in the Network,
ASONAM 2012, Intern. Conf on Advances in Social Network Analysis and Mining,
Istanbul, Turkey
August 2012.
-
[PT12b] J.L. Pfaltz
Entropy in Social Networks,
SocInfo 2012, Intern. Conf on Social Informatics ,
Lausanne, Switzerland,
December 2012 (submitted).