Word indexes are indexes that support implicit proximity between a provided
list of words.  String indexes are indexes that require exact matches on the
entire content of the supplied search term.

Bath uses structure 2 because they aren't requiring proximity searching from
their servers and plan on sending in single term searches.  If a Bath client
sends in a single term search to a Word index, they still won't get any
proximity searching.

Give me an AdjacencyWordList in Bib-1 and I'll use it.  I've only been
asking for it for 10 years.  No, I don't think we should use the one from
the Utility attribute set.  In the meantime, many servers do proximity
searching on WordList.


-----Original Message-----
From: Ray Denenberg [mailto:[log in to unmask]]
Sent: Thursday, September 19, 2002 9:47 AM
To: [log in to unmask]
Subject: Re: DC Index definitions

First, what do we mean by a word index?  In Bath we mean it for searching a
single word (Bath uses structure 2). Don't we mean the same for DC?  Let's
be sure we agree on this one way or the other.

Second, the alternative mappings proposed are (1) old architecture (e.g.
bib-1) and (2) new architecture (xd, utility, bib-2). Structure value 2 is
proposed only for the old.  AdjacenyWordList is in the Utility set (new
architecture). You can argue that new-architecture-friendly sets can be used
with old architecture, if not vice versa, but do we want to do that?

Finally   --- WordList????  Haven't we depricated that completely out of


"LeVan,Ralph" wrote:

 I'm loading this stuff into a file that I can process and I noticed that
the word indexes have a Structure attribute of Word (2).  Shouldn't that be
out new AdjacencyWordList, or at least just plain WordList(6)?Thanks!Ralph