[Tagdb] RDBMS, Lucene or both?

[ogjunk-tagdb at yahoo.com]

Ace,



My gut and experience with Simpy tells me your gut is headed in the right direction.  Two guts think alike.  RDBMS and Lucene are not exclusive and can go hand in hand.  There may be some data duplication, but in today's world of computing, this is not a big concern.

Otis
--
Simpy -- http://www.simpy.com/ -- Tag.  Search.  Share.


----- Original Message ----

From: Ace Jayz <fourtlove at gmail.com>

To: tagdb at lists.tagschema.com

Sent: Tuesday, February 6, 2007 1:43:15 PM

Subject: Re: [Tagdb] RDBMS, Lucene or both?





And perhaps most relevant to your question is the consideration that

the Solr team is paying to tagging, with use case and implementation 

possibilities being brainstormed here: <http://wiki.apache.org/solr/

UserTagDesign>.



        Erik







Erik,



Thanks for the pointer to the Solr Wiki discussion on tagging.  I've been thinking along similar lines and convinced myself that Lucene could efficiently handle many of the queries required of a tagging system.  I would like to avoid an RDBMS and all of its complexities if possible, but there are some operations that I think would be somewhat difficult with Lucene alone.  For example, gathering the data to build a tag cloud would seem to be problematic, without a lot of data duplication in the text index.   It boils down to trying to choose the right tool for the right job.  IMHO, Lucene is the right choice for the text search portion of a tagging system.  But an RDBMS seems like the right tool to represent the relationships between the tags, users, bookmarks, etc.  My one concern is that I go down the Lucene-only path and get most things to work well but that I fail to anticipate a future requirement that would be easily handled by an RDBMS and very difficult, if not impossible, with Lucene.  My gut tells me that some combination of the two is the right answer... 



-- Ace. _______________________________________________

Tagdb mailing list

Tagdb at lists.tagschema.com

http://lists.tagschema.com/mailman/listinfo/tagdb