stop words as blacklist

[John Blythe <jo...@curvolabs.com>]

hey all,

is there any out of the box way to use your stop words to completely skip a
document? if something has X in its description when being indexed i just
want to ignore it altogether /  when something is searched with X then go
ahead and automatically return 0 results. quick context: using solr for
semantic matching, sloppy ML'ing of sorts, so when running so automations
to connect incoming data to our own corpus of known items there are
irrelevant things that i'd like to not just block the stopword from being
part of the token stream, but block the entire document/search.

thanks for any info!

Re: stop words as blacklist

[Binoy Dalal <bi...@gmail.com>]

Like Ahmet says, a custom update request processor is the best way to go,
and it's pretty simple too.
I have a ready to use example here:
https://github.com/lttazz99/SolrPluginsExamples

On Fri, Mar 18, 2016 at 9:21 PM Ahmet Arslan <io...@yahoo.com.invalid>
wrote:

> Hi John,
>
> Do you want to skip that document in the indexing process?
>
> Or, you want to index that document, but you don't want to retrieve it if
> it is queried with stop words?
>
> There is a KeepWordFilterFactory to detect if a document contains a
> black-list word.
>
> To skip a certain document that meets certain criteria, Update Request
> Processors is the right place to look at.
> https://cwiki.apache.org/confluence/display/solr/Update+Request+Processors
>
> Ahmet
>
>
> On Friday, March 18, 2016 4:44 PM, John Blythe <jo...@curvolabs.com> wrote:
> hey all,
>
> is there any out of the box way to use your stop words to completely skip a
> document? if something has X in its description when being indexed i just
> want to ignore it altogether /  when something is searched with X then go
> ahead and automatically return 0 results. quick context: using solr for
> semantic matching, sloppy ML'ing of sorts, so when running so automations
> to connect incoming data to our own corpus of known items there are
> irrelevant things that i'd like to not just block the stopword from being
> part of the token stream, but block the entire document/search.
>
> thanks for any info!
>
-- 
Regards,
Binoy Dalal

Re: stop words as blacklist

[Ahmet Arslan <io...@yahoo.com.INVALID>]

Hi John,

Do you want to skip that document in the indexing process?

Or, you want to index that document, but you don't want to retrieve it if it is queried with stop words?

There is a KeepWordFilterFactory to detect if a document contains a black-list word.

To skip a certain document that meets certain criteria, Update Request Processors is the right place to look at. https://cwiki.apache.org/confluence/display/solr/Update+Request+Processors
 
Ahmet


On Friday, March 18, 2016 4:44 PM, John Blythe <jo...@curvolabs.com> wrote:
hey all,

is there any out of the box way to use your stop words to completely skip a
document? if something has X in its description when being indexed i just
want to ignore it altogether /  when something is searched with X then go
ahead and automatically return 0 results. quick context: using solr for
semantic matching, sloppy ML'ing of sorts, so when running so automations
to connect incoming data to our own corpus of known items there are
irrelevant things that i'd like to not just block the stopword from being
part of the token stream, but block the entire document/search.

thanks for any info!