Re. Indexing

[Geoff Hart <geoff-h -at- MTL -dot- FERIC -dot- CA>]

     A few folks asked for "true confessions" about how we index. I confess
     to being a traditionalist... I do it entirely by hand. (Yes, this
     _will_ be a problem for my 1000-page magnum opus.)

     I use the philosophy that the ultimate index contains at least one
     entry per paragraph: since each paragraph contains a single thought
     (or a few closely tied thoughts), it makes sense to provide an index
     entry to point readers towards that thought. (Think of this as a
     context-sensitive index.) This approach would be completely
     unmanagable under really tight deadline pressure; in such cases, I
     resort to examining each section (heading, subheading, and
     sub-subheading) and coming up with three or four keywords that
     describe the section.

     Finally, I do duplicate entries. As a user, I really hate entries that
     say "ROTFL: see rolling on the floor laughing", particularly when
     there's a single page number after the cross-referenced text. Note:
     this poses the problem of what to do when ROTFL doesn't appear in the
     text but the spelled-out phrase does. Here, I like to use "rolling....
     ("ROTFL" in text): page 123".

     Could I get someone to pay me for the countless hours this would take
     to do right? Generally no. But it's the only way I'm aware of to
     produce a really useful index that is adequately context-sensitive.
     New tools are coming out that use fuzzy logic technology to add
     context, and these appear promising.


     --Geoff Hart    #8^{)}
     geoff-h -at- mtl -dot- feric -dot- ca

     Disclaimer: If I didn't commit it in print in one of
     our reports, it don't represent FERIC's opinion.