Finding matched terms

[Walt Stoneburner <wa...@gmail.com>]

When performing a query and getting a result set back, if one wants to
know which terms from the query actually matched, is Highlighter still
the best way to go with the latest Lucene, or should I start looking
at query term frequency vectors?

Just trying to find a non-expensive way of doing this without using
the explain() method as shown by the Lucene in Action book, which
appears to incur at least a full order of magnitude performance
penalty.

Searching http://www.gossamer-threads.com/lists/lucene/java-user/
seems to suggest the Highlighter.

What makes this interesting is that, technically, I don't need to know
where, nor how many times, a term was found in the resulting document,
only that it occurred.  Any way to optimize for this special case?

e.g., CAT OR DOG
animals.txt -- CAT, DOG
felines.txt -- CAT
canines.txt - DOG

Thanks,
-wls

---------------------------------------------------------------------
To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
For additional commands, e-mail: java-user-help@lucene.apache.org