| The
Goal |
Code
Zebra seeks to build a cross-media dialogue generation and
analysis software machine, synchronous and asynchronous, networked
but viable on location, a critical tool, a moderator's tool,
able to map and mine data on site, in its archive and on the
web; visualization of relationships, dynamics, time relationships,
persistence, threaded to other related or opposition positions,
attract, able to map and remap, able to suggest relationships,
able to analyze and visualize data bases, beauty, individual
profile/group profile, facilitate affinities, visualization,
reaction diffusions patterns and other related natural sets,
evolutionary, "Dialogue Styling", play, able to work with
streams, arbitrary values, biases, networks, shared states,
disruptions, resocialize users, suggest new relationships,
disrupt dialogue patterns. Peace keeping software. "Visualize
the dynamics power of information gathering"; from highly
structured hierarchical relations (actors with scripts, moderated
environments, conferences, to more dynamic), know how YOU
like to engage and shop for the feeling of the conversation
as much as the topic; analyze structure or give the structure
its own intelligence; offshoots of group can find own I.D.,
keep relationship to rhizome. A major hypothesis of the project
is that we can use the patterns found on animals - strips,
spots, etc. - to help the user identify themselves in such
a space, and to give a recognizable identity to the space
as whole.
The Code
Zebra residency was undertaking in order to articulate and
outline a software that embodies chat and discourse analysis
on line, considers live events and the affective qualities
of these and enables long term focussed dialogues between
individuals and groups from diverse cross disciplinary communities.
The software seeks to facilitate the performance of identities,
in the now time span of the Internet and streamed media and
in the historical sense of a topical digital archive. It will
use meta data analysis, text analysis, and computational linguistics
and performance theory as well as advanced visualization techniques.
The software includes character sets that are identified with
specific familiar patterns and linked to ways of acting. While
the software allows different patterns to emerge through the
chat and discourse analysis, there are a number of these characters
embodies within the software which activate when participants
begin to talk in certain ways.
top
|
|
Major
Issues
|
Individual
vs. Global Patterning
If scientists
are giraffes (cautious, always moving), artists ocelots (small,
mischevious, energetic) and critics leopards (large, waiting
and watching and then pouncing and dominating), how do we
allow people to distinguish their visual identity within these
groupings? And if they are all aggregated together, will they
form a global pattern which is recognizable and meaningful?
Pattern
vs. Information vs. Meaning
Zebra patterns are recognizable, but are they meaningful to
us? Do differences in the pattern help us distinguish differences
in quality, or heredity, or disposition? If we use such patterns
in a software system to identify individuals, types of conversations,
and the general state of the entire conversational topology,
will users be able to read them?
Qualities
of Conversations
What information do we want to extract from the conversations
which are ocurring? Rate of reply, length of post, frequency
of post; enflamed, reflective, meditative, aggressive, defensive;
enouraging, supportive, dismissive, etc...
Migrating
Through Patterns
Are the patterns fully determined by automatic processes or
by the users themselves? And how do the users' patterns change
over time, and how then does the patterning of the entire
space change over time?
Starting
From Zero
How do we bootstrap the envirnoment into a state which contains
meaningful information?
What's
the Maths Involved?
Can Voronoi diagrams be employed? What about reaction-diffusion
patterns?
top
|
| Convergence |
Over the
course of 10 days, we boiled the major issues down to three
high-level questions pertaining to making the software concrete:
How
to represent large-scale chat/dialogue spaces?
In other words, what visual language is appropriate for conveying
the dynamics? Reaction - Diffusion, Voronoi...
How
to allow user to simultaneously participate in a particular
thread while attending to the larger conversational geography?
2-D space with 3-D pockets, personal disks
How
to create patterns which are true to the data, meaningful
to the user and maintain the animal-skin aesthetic?
top
|
