Stories and social networks - Narrative Intelligence - page 309

attempts to design and implement computer programs that combine sophisti-
cated computational linguistic analysis with social network analysis are as yet
unrealized.
Te chno l og y
SystemDesign and Implementation. I have been analyzing Usenet newsgroup,
audience discussions of popular television programs in an attempt to under-
stand how the stories of television are pulled apart, reiterated, quoted, summa-
rized, and - in general - appropriated into and used for the social networks of
television viewers.
To analyze these and other newsgroups the Conversation Map system has
been designed and implemented. The input to the system is an archive of thou-
sands of messages from a newsgroup. The output of the system is four-fold and
is pictured in the figure below.
1. Social Networks : The upper left-hand panel displays a social network show-
ing who is in conversation with whom. The nodes of the network are labeled
with the names of the participants in the newsgroup conversation. If two names
are connected and close to one another, then the two participants have been
responding to or quoting from each other more frequently than if they are
connected but far apart from one another. Two names are connected if both
participants have responded to or quoted from the other. In other words, the
social network diagrams reciprocity . If someone in the conversation posts a lot
Figure 1. The Conversation Map interface.