You can go the other way and prioritize

[hrsibar4405]

In the search results we get “Red Fairy Dress”, because “shoes” can be not only in the name of the product being searched for, but also in the description of other products.



In this case, you can adjust the query and search only in the title, but...



... we have not yet seen a store where you kazakhstan telegram number database would need to search only by the name of the product, and not by the description, value of the properties, etc.

(weight) the fields.



This allows for more flexible ranking control depending on the field in which matches were found.

The search results also depend on the analyzer used, which converts the original text into a structured format optimized for efficient storage and fast search. ES has several built-in analyzers, and their results differ.

Comparison of analyzers.png

It is not always possible to get what you need from standard analyzers. In this case, you can make your own by modifying one of its elements - a tokenizer. There are several available tokenizers, for example, N-gram tokenizer or Edge-ngram tokenizer. The latter is usually used to analyze queries as you type.