I would like to store multilingual documents, perform full text search on all documents in all languages, and in search results if a match is found in multiple languages for the same document, I would like to keep only the most relevant language.
To store documents in multiple languages there is already multiple posts regarding this topic (Multilingual elasticsearch indexing best practice/experiences): - option1: one index by language - option2: multi_field/separate fields
So with option1 the search result could contains: - doc1_en - doc1_fr
How to remove duplicates ?
With option2 the search result will contains only one result but how to identify the most relevant language ?