Try setting the slop factor on your phrase query. This should accomplish what you want. Set it to something like 10 and see what you get.

Erik


On Mar 16, 2004, at 8:55 PM, Supun Edirisinghe wrote:


I have a field called buisnessname and this field contains keywords like
"Georgian House" "Georgian" "The Georgian House Hotel" "Georgian blah
blee bloo Hotel" along with 10,000s of other documents that have the
word 'Hotel' somewhere in the businessname field.


When I do a phrase query on "Georgian Hotel" I get only the one document
back. I would like to get that one back as the top result but also the
other stuff that has "Georgian" and "Hotel" too. Also, I'd like to have
"Georgian House Hotel" show up before "Georgian blah blee bloo Hotel"


Right now I do an or'd boolean queary with
each of the words in the the search string as a Term in business name
as well as
the entire search string as an exact PhraseQuery and boost that by 3.

But this doesn't allow me to ensure that "The Georgian House Hotel" will
come before "Georgian blah blee bloo Hotel". (there are other fields
queried besides business name) and in my instance of the index,
"Georgian blah blee bloo Hotel" comes out with a better score because of
other fields). I would like the the closeness of the phrase to be taken
into account. any ideas on constructing a good query for this situation?



thanks



--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



  • phrases Supun Edirisinghe
    • Erik Hatcher

Reply via email to