thinking analogously in the sense that re-presenting large sets of
data is going to require an intensely rich grammar for the meaning
laden/pregnant manipulation and agglutination of these objects about which we
have been talking.
If I
am someone who is interacting with large biomedical data sets, I may need a set
of primitive objects each of which can represent some elemental of the vast
data field. For example, let's say I am
a user who is responding to a flood of sensor data which has suddenly detected
germs of some kind following an explosion on a battlefield, or in a subway
terminal--take your pick. So off go my
sensors which are distributed all around the area of detonation. As the data starts to come into the Grok Box
interface environment, below is a rough picture of the kind of thing I am
fantasizing I would see in an array of screens spread out before me as the
expert medical/military bio-counter terrorism response dude. The fantasy scenario would unfold in the
following order. BTW, this is pure
scientific friction. This is what I
would want to see were I in such a situation.
- a single, simple object
which appeared soon after the explosion hovers in a large screen directly
in front of me. Let's say this
single prominent object is a simple circle, white in color. This 2 dimensional object is suspended
against a black background as a heavy stream of live data continues to
flow unseen into the system and I wait for what to do next.
- Suddenly, the object
begins to three dimensionalize—fill out into 3 space and following the
filling out quickly turns from white into a very intense green color. As the bio expert dude, I know that
this means the explosion has released not a toxic chemical, in
which case the sphere would have turned into a black color, but has released
a germ the identity of which I have no clue about yet.
- The object now hovers in
its new state in front of me
- Next, I reach my virtual
hand out and touch the sphere object (had there been more than one germ or
chemical, the system would have put two, three, or four objects in front
of me). By clicking the one, or a
particular one, the machine knows to begin the process of identifying that
germ/chemical represented by the object.
- The sphere is now going to
change/morph in ways which will tell me whether the organism is a virus or
a bacteria. For example, if a
virus, the shape may become like that of a 10 or 20 (icosahedral) sided
dice; or else, the object may change into a shape more akin to a spiral. If the organism is a bacteria, the
object may change into the rough shapes of a paramecium or an ameoba. Perceptually/culturally, such shapes
have been associated with the two kinds of organisms mentioned and so are
reasonable first iterations of how we may visualize that streaming sense data
which intially identifies the kind of germ we are dealing with.
- Once the object assumes
one of these shapes (in this case, the organism is a virus as indicated by
the sphere changing into a 20 sided shape), I reach out again and touch
it. At this point, the object
begins to open and out of it comes an array of say 6-12 squares which all fill my
screen along the x axis. So from
left to right along the bottom of the screen are laid out these square
objects. The background is still
black.
- Next, I see distinct,
individual, simple geometric forms appearing some distance above each
rectangle. These are initially
(perhaps) 3 dimensional and white as they come out of the square and hover
above it until the data begins to stream into and morph the object. By this process, I know the machine has
begun the biochemical analysis of the germ’s elemental components
(Topper…each of these objects above the rectangles represent the
“primitives” I need you to create)
- The still objects above
the squares, after a time, will begin to change in accord with the
specific data that comes into them.
That is, data from sensors, and soon from queried database stores,
are going to be effectively streaming right into the individual objects,
thus structuring them--loading them with meanings/significance in
real time with real data.
Presuming the ground sensors to have sufficient
biologic/biochemical sensitivity, as the user I would like to see the
first object on the left representing perhaps parts of the amino-acid/protein
schema of the viral particle.
- Once the object has been
morphed by the data, I will then take that molecular/structural
information and instruct the system in such a way that an array of qeuries
are sent out over the network/web to begin the intensive database
searching which will result in the identification of critical features of
a suspected weaponized virus.
- Let’s say for the sake of
simplicity that there are 10 individual databases distributed throughout
the US which contain information relevant to the identification and neutralization
of the virus in question (its likely there would be many more than
this!). So in response to my
queries, data from these 10 data bases begins to stream back. Critical at this stage in the processs
of visualizing all of this data to myself, for the sake of the example, is
to connect each data base with one of the objects above the squares. These objects may themselves become
multiple objects once the data starts getting represented.
- Next, the individual
objects and their associated squares spread out into their own space at
the outer sections of the very large screen I am interacting with. The center of the screen remains
empty. In their respective
locations the initial objects begin to break up into multiple distinct
objects which are now undergoing intensive alterations and enrichments so
as to represent the raw data coming in from their respective database.
- For example, several
groups of objects are taking data from sources dealing with the capsids of
possible viruses (ie the viral protein coat). As the identity of a virus
will depend a great deal on the structure of its capsid, each object in
this group may be representing properties, morphologically and abstractly
(not necessarily in terms of mirroring the shape of a certain molecule),
of different parts of the viral protein coat.
- Another group may be
taking data from a source dealing entirely with content related to RNA
virsuses in general
- Another group may be taking data from a source dealing
entirely with content related to DNA virsuses in general
- As the system learns more
about the germ it is dealing with it refines its search parameters.
- A set of objects may then
be visualizatin data from a data base whose sole content is recent
research on a family of RNA viruses known as Filoviridae.
- as the filtering and
visualization are going on the second to final stage in the process of
identifying the virus is intitiated.
This involves particular data object shapes from each of the groups
either
·
bonding in specific ways to form compound objects; this
bonding/fusing occurs according to the morphic/textural/chromic/dynamic rules
of “grammar” which the system knows to apply to the abastract objects for the
particular content set being dealt with (ie, molecular microbiology,
biochemistry, pathophysiology, immunology, and pharmacology).
·
Staying singular.
- the final stage of the
metavisualization which will yield the identity of the entire virus, or
else key antigenic regions of it for the targeting of drug and/or immune
actions is now primed and ready for execution.
- in this final stage I as
the user am watching all these grouped objects at the periphery of the
screen space. Some of the groups are going to have information absolutely
critical and others will have information at a 2nd or 3rd
layer of importance. The system
will alert me to the simple or compound objects to which I must
attend.
- Because of my own
specialized training with this technology, I have understandings of what
is going on such that I, and not the machine, have to make the final
decisions as to how and in what order the relevant objects are to be
grouped together. So I begin the
final stage by dragging the prepared data objects from the perimeter and
placing them in the center of the screen within specific regions.
- With each placement, the
compound object grows and complexifies.
- The process of
‘agglutination’ of the various peripheral objects has resulted in a highly
significance rich complex entity which provides me information as to the
idenity of the virus in question: an RNA virus, Family: filoviridae;
Genus: unknown; Species: Ebola-Zaire; Special Features:
altered (ie weaponised) into aerosol.