Item14806: Solr too much unrelated results

pencil
Priority: Normal
Current State: Closed
Released In: n/a
Target Release:
Applies To: Extension
Component: SolrPlugin
Branches: master
Reported By: UlrichLeodolter
Waiting For:
Last Change By: MichaelDaum
I got many complaints about too much unrelated results and i cannot find what i am looking for

We are running the lastet Foswiki 2.1.6 + SolrPlugin + solr-5.5.5. I tried to understand the reason behind that complaints und found text_prefix and text_suffix the most problematic one. When searching e.g for Leodolter, i don't expect leod, leodo, leodol, leodolt or leodolte in my results.

Especially the facets are not useful when you got more than 100 results, but only 5 contain the terms you searched for.

I tried to exlcude text_prefix, text_suffix at search time in solrconfig.xml, that seems to work.

  text_prefix^0.5
  text_suffix^0.5

Wildcard search seems also work, are indexed fields text_prefix and text_suffix used during wildcard search?

There is an interesting blog entry on this topic: http://blog.florian-hopf.de/2014/03/prefix-and-suffix-matches-in-solr.html

I tend to disable ngram at search time and instruct users to use wildcard if they are unsure.

-- UlrichLeodolter - 27 Dec 2018

This was fixded by using different parameters during analysis and indexing time to prevent search strings from being split up and injected into the query.

-- MichaelDaum - 31 Jan 2019
 
Topic revision: r2 - 31 Jan 2019, MichaelDaum
The copyright of the content on this website is held by the contributing authors, except where stated elsewhere. See Copyright Statement. Creative Commons License    Legal Imprint    Privacy Policy