Being worked on by [Pourang Irani]
i emailed PourangIrani?, and he wasn't aware of one; but i am not conversant in the field of InformationVisualization, so i can't confirm independently. But i would think he would know. -- BayleShanks
I'm looking into it. Getting closer.
Later... Pourang co-published a paper on Geons with ColinWare?, which suggests that I might have seen the diagram in Colin's book. I'll check that when I get home. --ss
Here it is: [WilmaScope]: Wilma was originally created by Tim Dwyer and Peter Eckersley. Its a very interesting and beautiful program written in java, but it soaked up too many resources on the linux and windows and organic ;-> computers I tried it on, which led me to begin a visual python 3D force-directed layout program of my own (incomplete, and not even beginning to deal with UML)
sorry, i forgot to define geons. geons are a class of 24 simple 3D primitives that are probably particularly easy for a visual system to recognize when projected into 2D at an arbitrary viewpoint (they are almost "viewpoint-invariant"). ""The geons have two particularly desirable properties: they can be distinguished from each other from almost any viewpoint, and their identification is highly resistant to visual noise."
"The set of geons is defined sot hat they can be differentiated on the basis of dichotomous or trichotomous contrasts of viewpoint-invariant properties to produce twenty-four types of geons." That is, each geon is defined by a set of binary and trinary feature values, where each feature is an invariant property that can be detected in an image. This measn that you can go from detected image features to specification of a geon. "Deriving the geons from contrasts in viewpoint-invariant properties renders the geons themselves largely invariant under changes in viewpoint."" (from [here])
I can't quickly find pictures of all 24 of them, but here's some:
So some people hypothesize that geons are the primitives in human vision, and hence should be especially easy to recognize or remember.
So a UML --> GeonDiagram tool would allow specification of the icons of the nodes in terms of geons or the composition of geon primitives.
I would assume this functionality could be easily added onto a tool like [WilmaScope].