Danyel Fisher
Toward Articulating Social Roles in Online Spaces
 |
Contact Information
Microsoft Research
|
Biography
Danyel Fisher joined MSR in November 2004, when he started a
course of research centering on online behavior centering on Usenet newsgroups.
His disciplinary background has applied social network analysis to a variety of
online spaces, centering on email archives. He earned his PhD in 2004 from the
Informatics department at UC Irvine, working with Paul Dourish.
Position Paper
Community organizers are very aware of the different sorts
of roles that seem to appear in online spaces: a long bibliography of work
about online community talks about roles like “central contributors”, “flamers”,
and “lurkers.” While these labels are easy to attach on an ad-hoc basis, there
has been little work attempting to systematize what sorts of people occur in
different online spaces, or how their interactions might be matched to these
roles. My recent research at Microsoft’s Community Technologies Group attempts
to use structural and social network information—social metadata—about messages
in Usenet Newsgroups to help identify both the sorts of roles that occur in a
space, and to try to understand what is the shape of the space around these
people.
There are obvious values to this information to community
organizers, who have good reason to want to know who is in their space and what
they are doing; there are longer-term implications for providing feedback to
users that helps give them more information about to whom they are talking and
what they might want. Imagine a group interface, for example, that could give
information about whether a group is filled with discussion, consumed by flame
wars, or centered on questions and answers.
There are other, troubling implications to this work: this
is a finer-grained version of reputation, with much richer information about
behavior than the standard good/bad (e.g. eBay)’s reputation. There are
interesting challenges to what it means when an online system actually
remembers that a person likes to be involved in flame wars: like the boy who
cried wolf, will that person now be unable to get a straight answer?
Back to Social Computing Symposium 2005
|