Hello,
I'm new to Lucene. After some readings, I'm still not quite sure which
Analyzer I should be using for handling documents in unknown or multiple
lanugages. The documents I want to index may be written in lanuages other
than the user/system's default language and one document may contain t
: Could someone please give me some suggestions on how to implement date
: boosts? I would like to boost the document when it is new and lower
: the boost when it's old.
you should check out this older thread...
http://mail-archives.apache.org/mod_mbox/lucene-java-user/200501.mbox/[EMAIL
PROTEC
: >>[Query]
: >>"Napol* Dynamite" near "film|movie"
: >This can be done using nested SpanNearQuery's and SpanOrQuery's.
: >A PhrasePrefixQuery can not be used as a SpanQuery.
I've never really looked at SpanQueries very hard, but this thread got me
a bit curious.
Looking over the docs and the c
Hi
Could someone please give me some suggestions on how to implement date
boosts? I would like to boost the document when it is new and lower
the boost when it's old.
Thanks,
Ben
-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For a
Paul,
Thanks for pointing me to the surround code. I have started playing
with it, and am impressed. Now I just need to adjust my thinking a bit
more to see if I can implement the tool correctly, and get my specific
search functionality out of what it offers.
You've been a great help,
Sean
Chris,
I think my hesitation with my approach was due to being lazy, and
not wanting to get up to speed with ant. I've gotten over much of the
Eclipse internal project/build learning curve, so this is very likely a
case of the 'golden hammer' syndrome, where I am ignoring the benefits
of us
: For example, given this data:
:
: author: a b c
: author: d e f
:
: a search for "a SAME c" would match the first row, but "a SAME d" would
: match nothing, which is what I want.
if i understand you correctly, then you are describing a use case in which
the index has two documents, each contain
I don't use Eclipse, (and in fac i've never acctaully built the from
source) but if i remember correctly, one of the main reasons why the
"sandbox" was retired and everything in it was moved to where it is in the
"contrib" directory was so Lucene and all of the "contrib"uted code could
be compiled
Is there a way to tell Lucene to restrict proximity searches to just one field?
This would mimic the BRS/Search SAME-operator, which I use very often.
For example, given this data:
author: a b c
author: d e f
a search for "a SAME c" would match the first row, but "a SAME d" would match
nothing
Hello,
I am new to subversion, junit and the Lucene contrib repository. I
am looking over the 'surround' project at the moment. If there is anyone
out there with Eclipse experience who uses the contrib subversion (or
cvs) repository could you look over my approach listed below?
I am using
Hi Jeff,
This is a tough question to answer, because there is no universal
answer. The choice of Analyzer depends on what/how you are trying to
index/search. I've used analyzers from the Lucene distributions, but
have also written specialized ones. My suggestion for you is to start
with the Sta
Sean,
On Sunday 04 September 2005 20:43, Sean O'Connor wrote:
> Hello,
> I am trying to do some complex queries such as:
>
> [Field contents]
> The movie Napoleon Dynamite is a movie about a kid named Napoleon who
> has no Dynamite.
>
> [Query]
> "Napol* Dynamite" near "film|movie"
>
>
>
I believe the index just contains information about single terms. A
PhraseQuery then searches the index for the parts of the phrase and
returns the hit information.
So, as far as I understand, there is no way to get the frequency of
phrase directly from an index, but you could create a PhraseQ
Hello,
I am trying to do some complex queries such as:
[Field contents]
The movie Napoleon Dynamite is a movie about a kid named Napoleon who
has no Dynamite.
[Query]
"Napol* Dynamite" near "film|movie"
Is this possible with some version of a span query? Something like a
PhrasePrefixQ
"MMDD" is better since it has less number unique terms compared with the
unix time stamp if you only care about the days.
-Original Message-
From: Filip Anselm [mailto:[EMAIL PROTECTED]
Sent: Sunday, September 04, 2005 3:56 AM
To: java-user@lucene.apache.org
Subject: Re: How to searc
Question to those who've deployed and maintained Lucene: any recommendations
or observations about practical decisions regarding analyzer choice in
indexing & searching? What have you found in operation to work well, become
difficult, yield better/worse results, affect performance, etc.? What wo
DateFilter sounds great!! - But how is the best way to store dates in af
Field? I get the time as a unix time stamp, seconds since epoch - and
usually I can cut it down to hours or days since ephoc instead - if this
has any effect on the perfomance...
thanks...
Chris Hostetter wrote:
>: How do I
: How do I combine two queries - one made by the QueryParser and the
: programmatically made RangeQuery?
you could make them both children of a single BooleanQuery, but as long as
you're going to write a little java code to put them together -- why not
use a DateFilter instead?
http://lucene.apa
jian => I'll try the RangeQuery first, and if it doesn't give any
perfomance problems I'll just stick to that method - it's easy, simple
and doens't involve other systems. But thanks for info...
How do I combine two queries - one made by the QueryParser and the
programmatically made RangeQuery?
F
19 matches
Mail list logo