lucene-java-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Mathieu Lecarme <math...@garambrogne.net>
Subject Re: Synonym Search and stemming
Date Mon, 23 Jul 2007 13:45:55 GMT
IMHO, stemming is hurting index. A stemmed index can't be use for
completion or other kind of search.
But stemming is nice for synonyms search. You should look for
spellchecker code. If you index your word with stemmed version, you can
provide a synonym filter, just like wordnet example.

M.

Le lundi 23 juillet 2007 à 18:27 +0530, sandeep chawla a écrit :
> Hi,
> 
> I want to implement synonym search in my application.
> 
> I think there are two ways to implement synonym search.
> Index the documents using a synonym analyzer
> OR
> Expand the query using a parser which uses a synonym analyzer.
> 
> Well  I am expanding query as indexing with synonym analyzer takes
> much time and more space too.
> 
> Also I am using PorterStemmer in indexing.
> 
> Now suppose i want to search for word "simple"
> 
> There are two ways now..
> 1----
> 
> First i stem this word to "simp" so that i can match all the variants
> of "simple" in my index like simply etc.
> 
> then i expend this query using synonym analyzer but it doesnt get
> expand as i m using the stemmed word for synonyms..ie. "simpl".
> 
> so when query.toString
> 
> It is field:simpl
> 
> I just get all the documents which have only "simple"
> 
> 2----
> 
> First expand the query using synonym analyzer
> then i stem it.
> 
> In this case
> 
> Query.toString = field:"simpl elementari uncompl childlik unproblemat
> unsubdivid"
> 
> I would like it to  be
> 
> field:simple field:elementari field:uncompl field:childlik and so on
> 
> i'd appreciate any help in this regard
> 
> Thanks
> Sandeep


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


Mime
View raw message