This module is for advanced customization of the Elasticsearch Indexer and Search Engine for Magento 2.3. Unless you are familiar with Elasticsearch, you should discuss any changes with the CPCommerce team.
Enabled - Yes/No, Default No
Enable or Disable all functionality
Number of Shards - Number between 1 and 10, Default 5
Number of shards for each store index.
Higher numbers of shards increase performance/speed of search while decreasing the accuracy of the results.
Lower numbers of shards are suitable for smaller catalogs where search relevance is the most important aspect.
How many shards are appropriate
Default Tokenizer - Default 'standard'
The standard tokenizer will break apart a string based on English grammar, breaks on whitespace, hyphens, punctuation, etc.
Standard tokenizer can cause problems with things like hyphens in SKUs, so you can change to whitespace only in that case.
Synonyms allow you to enter various phrases that should all be considered identical in indexing and search.
E.g. if your data has "ipod" and someone searches for "i pod" or "i-pod" you can add a synonym to make sure those items are scored as if the end user typed the correct spelling/phrasing that you have in your data.
Synonym Token Filter Reference
By default, all string data is broken apart based on the tokenizer above. This limits the ability to find partial matches in your data. Ngram attributes are the attributes you want to allow partial matches for.
E.g. If your data contains an item named "Yellow Shirt" and you want someone to be able to find that item by search for "yell shir" you would need to use ngram attributes for that search to match.
This is also useful for partial sku matches if your sku conventions contain some grouping mechanism
E.g. A sku of APPR1230 or APPR3010, you would use ngram on SKU to allow someone to search for APPR and find apparel skus. You could combine this with synonyms to have APPR be a synonym for apparel, which would allow APPR skus to match a search for apparel.