Edge-NGram Ranking Results Incorrectly?

Elastic Stack Elasticsearch
1

I noticed the edge_ngram analyzer scoring exact matches of a query lower than poor matches so I copied the data into a second field with the standard analyzer to compare.

I now have 2 fields with the same content but different analyzers:

  • text_xx: standard analyzer
  • text_en: edge_ngram analyzer (3 to 20 chars)

Searching on the edge_ngram field yields the wrong rank/score IMHO.

I have a document which has text_en/text_xx="UVV 00031 099 Filterschlauch" and when I search for "UVV 00031 099" I expect this document to appear first. There are 50K+ documents containing "UVV".

  1. Searching (UVV 00031 099) I get the result I want in second position:
    image

  2. Searching text_en: (UVV 00031 099) I don't even get the document at all:
    image

  3. Searching text_xx: (UVV 00031 099) I get the correct document in first place:
    image

Why does the Edge-NGram analyzer not score the exact match highest?

See the explanation of the search here.

2

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.