Wrong facets for multi word queries

[fsteeg]

http://lobid.org/resources/search?q=ehrenfeld+nubbel

Facets shown would be correct for just querying 'ehrenfeld'.

[fsteeg]

It's actually not the counts that are wrong, but the searches linked from the facets.

[fsteeg]

Deployed to stage: http://stage.lobid.org/resources/search?q=ehrenfeld+nubbel

[ChristophEwertowski]

Doesn't work for libraries: For example, if you select the facet "Köln USB" after executing the search above you get 298 hits while the facet counts 334 (staging and productive).
For the rest: +1

[acka47]

Doesn't work for libraries: For example, if you select the facet "Köln USB" after executing the search above you get 298 hits while the facet counts 334

I think we already had this in the past. Isn't the underlying mechanism that the facet gives the number of items while the query result lists the FRBR manifestations (or in bibframe-speak: instances)? This would mean there are 36 dublicate items...

A good example is the setting the holding facet on "Bibliothek des Ruhrgebiets". You get only three examples but the facet count says "44". When you take a look at the resource "Die Ehrenfelder" you see that a lot of items are connected with it.

[acka47]

@fsteeg I don't know how this count is implemented? Could you just count distinct entries in the itemOf field of the items in question?

[ChristophEwertowski]

Shouldn't we have the discussion why it doesn't work in #207 ?

[fsteeg]

I'll look into the owner issue, but since it is unrelated to this issue (which affects all facets, and much more severly) I second @ChristophEwertowski's suggestion to discuss this in #207. Will reopen that.

[acka47]

+1