4. Organization:
This is an area where the current Net has some of its most visible
problems, as we have all struggled through myriad gopher menus, ftp
sites, and web pages looking for something that we know is there but
cannot find.

There is no ideal organization of information, but no organization is no
ideal either.  The organization that exists today in terms of finding
tools is an attempt to impose order over an unorganized body.  The human
mind in its information seeking behavior is a much more complex question
than can be answered with a keyword search in an unorganized information
universe.  When we were limited to card catalogs and the placement of
physical items on shelves, we essentially had to choose only one way to
organize our information.  Computer systems should allow us to create a
multiplicity of organization schemes for the same information, from
traditional classification, that relies on hierarchies and categories,
to faceted schemes, relevance ranking and feedback, etc.

Unfortunately, documents do not define themselves.  The idea of doing
WAIS-type keyword searching on the vast store of textual documents on
the Internet is a folly.  Years of study of term frequency, co-
occurrence and other statistical techniques have proven that keyword
searching is a passable solution for some disciplines with highly
specific vocabularies and nearly useless in all others.  And, of course,
the real trick  is to match the vocaubulary of the seeker of information
with that of the information resource.  Keyword searching not only
doesn't take into account different terms for the same concepts, it
doesn't take into account materials in other languages or different user
levels (i.e. searching for children will probably need to be different
than searching done by adults, and libraries actually use different
subject access schemes for childrens' materials).  And non-textual items
(software, graphics, sound) do not respond at all to keyword searching.

There is no magical, effortless way to create an organization for
information; at least today the best tools are a clearly defined
classification scheme and a human indexer.  At least a classification
scheme or indexing scheme gives the searcher a chance to develop a
rational strategy for searching.

The importance of organizational tools cannot be overstated.  What it
all comes down to is that if we can't find the information we need, it
doesn't matter if it exists or not.  If we don't find it, we don't
encounter it, then it isn't information. There are undoubtedly millions
of bytes of files on the Net that for all practical purposes are non-
existant .

My biggest fear in relation to the information highway is that
intellectual organization and access will be provided by the commercial
world as a value-added service.  So the materials will exist, even at an
affordable price, but it will cost real money to make use of the tools
that will make it possible for you to find the information you need.
If we don't provide these finding tools as part of the public resource,
then we aren't providing the information to the public.