lucene-java-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Erick Erickson" <erickerick...@gmail.com>
Subject Re: Unique City, State results from index based on zip
Date Wed, 04 Apr 2007 13:36:58 GMT
The default operator for QueryParser is OR, so what you may really
be getting is hits on Mill, and Vally is irrelevant.

But this is just a guess, it'd be way more helpful if you told us what
your index structure was and what query you actually submitted,
for which query.toString is really helpful.

And if you haven't gotten a copy of Luke, you should. It'll help
you examine your index, see how different queries parse (be sure
to set the Analyzer you use in Luke), etc.

Erick

On 4/3/07, freaktet <mlmyers@gmail.com> wrote:
>
>
> I am having the following problem.
> I have an index I built from a standard US Zip Code table.
> Users can search for any combination of City, State and Zip.
> If they search for City, State, I want to find unique results, instead of
> one result for every zip code that city state has.
> For instance, a search for "Mill Vally, CA" (mispelling on purpose), gives
> these results:
> 94941, Mill Valley, CA
> 29715, Fort Mill, SC
> 46365, Mill Creek, IN
> 29708, Fort Mill, SC
> 94114, Mill Valley, CA... etc.  Notice the duplicates.
>
> Any suggestions?
> --
> View this message in context:
> http://www.nabble.com/Unique-City%2C-State-results-from-index-based-on-zip-tf3523168.html#a9828966
> Sent from the Lucene - Java Users mailing list archive at Nabble.com.
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
> For additional commands, e-mail: java-user-help@lucene.apache.org
>
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message