Need help with Online Tables of Contents

["Doug, Data Librarian at Ext 4225" <engstromdd -at- PHIBRED -dot- COM>]

Folks:

I'm designing an online glossary of terms used in our "shared" data
models.  I need some help/opinions on designing the lowest level of table
of contents access.

We're planning to have two tables of contents.  One will be organized by
subject area (as defined by the data model), and will branch to the
entities included in that subject area.  For example:

      Location
      |  |  |
    Site |  Building
        Area

The other table of contents will be organized by major application, with
entries for payroll, product performance information system, general
ledger, etc.

In both cases, the next-lowest level may contain a very large number of
entries.  In the subject area index, an entity may have a huge number of
attributes, and a system may use a large number of terms.  The question
is, how do I organize the lowest level?  By logical groupings or
alphabetically?

For example, a grouping under "Site" might be "Position," which would
contain the attributes "Latitude," "Longitude," "Mailing Address," etc.  A
grouping under the "Payroll" program might be "Employee Information,"
which would include "Name," "Social Security Number" and so forth.

The danger here is that users won't mentally group the attributes or terms
the way I do, and once they make a wrong branch, it may be hard to get
back.

The other alternative is to organize all the terms for each entity or
application into a series of alphabetical lists.  The dangers here are that
users won't know how to spell the term, it duplicates the access I've
already provided with "Search,"  and it switches organizational schemes on
them in the middle of a search.

Does anybody out there (especially indexers) have experience or insight
with this type of problem?  Reply to me and I'll post a summary.

Doug
ENGSTROMDD -at- phibred -dot- com