picture lies in the term filter. If you want to go somewhere in some territory, you choose to consult a
map first rather than make SWAGs
[5]
and drive all over the place looking for what you want. That's
where topic maps come in. They are maps; only maps, and not the territory itself.
[6]
And maps, being
many things, are filters.
[5]
Scientific wild ass guesses.
[6]
The observation "The map is not the territory" has been attributed to Alfred Korzybski. See
http://www.gestalt.org/alfred.htm
for more information.
So, a topic map is just a map, and not the territory itself. How do I make a topic map more useful?
What does more useful mean? Now that's a focus question if I ever saw one. It seems to me that if you
want a map on which to plan the construction of, say, a new building, although you might start with a
road map used to navigate the town in which you plan the construction, you would proceed with a
topological map, perhaps one commissioned from a local surveyor. Thus I offer a response to the
"What does more useful mean?" question as follows: the map must represent the territory in such a
way that the application the map is intended to serve is best served.
[7]
You retort, "Say what?" to
which I respond that there is, indeed, a semiotic aspect to this discussion--the words need to fit the
problem space I have created. Let me explain.
[7]
I have always been a big fan of responses that don't say anything.
This book discusses the application of topic maps in the service of knowledge representation. That's
like uncovering an enormous snake pit.
[8]
First, there is the big question, "What is knowledge?" But
why are we considering that when I'm just trying to justify the claim that a map must represent the
territory in such a way as to be useful? I believe I am about to claim that a topic map is, indeed, a
member of the set of objects that intentionally represent knowledge. Heady stuff, that. A semiotic
stance dictates that we make sure that we do, indeed, represent that which needs to be represented.
Representing less would result in ample insufficiency, and representing more would result in
information overload. As my grandfather used to say, "You can't win for losing."
[8]
Have you ever uncovered a snake pit? Trust me, you don't want to go there, but we do it here
metaphorically anyway.
But as programmers we want to make sure we cover everything, which puts us at risk of generating
information overload.
[9]
How do we cover everything without swamping ourselves with too much
information? Hah! By making our topic maps more useful. Dang! Now we've gone full circle and
must reask the question, "What does more useful mean?" That, my friend, is what topic maps are all
about. Again, let me explain.
[9]
Notice that I still haven't answered the question, "What is knowledge?"
Topic maps are, indeed, automatically more useful--if done right. A topic map can be structured in
such a way that information that lies on a user's critical path can be presented directly while peripheral
information can be presented such that cognitive loads on the user are not increased by its presence.
Figuring out how to "do topic maps right" is the focus of this book.