whitehd at DREXEL.EDU
Thu Feb 21 16:30:59 EST 2008
I take it by "co-citation clusters" you mean what Garfield calls
Historiographs, that is, graphs showing linked papers over time.
That's what ch. 7 is about, and technically that's not co-citation.
If you can arrange your data to form a directed acyclic graph, so
that every paper points only backward toward earlier papers it cites,
you can use Pajek, available free on the Web, to visualize it.
See Vladimir Batagelj's paper, "Efficient Algorithms for Citation
Network Analysis" downloadable in pdf at:
Batagelj is the co-creator of Pajek, and there is a module for
certain kinds of citation analysis built into it. Note his
references to Hummon & Doreian's work as well.
More information about the SIGMETRICS