---
title: "Optimize keyword searches"
slug: "optimize-keyword-searches"
abstract: "How you can use synonyms and word-level relationships to optimize search results in Sitecore Search."
---
# Optimize keyword searches

Sitecore Search allows you to expand keyword searches by using semantic relationships between keywords or key phrases. With one-way or two-way synonyms, the keywords in a request can be related to other words, expanding the number of search results and keeping visitors engaged.

## Synonyms and replacements

In Sitecore Search, you can define one-way and two-way synonyms to create relationships between hyponyms and taxonyms.

> [!tip]
> There is a hyponym-taxonym relationship between the three words: *feline*, *lion*, and *tiger*.
>
> Because *feline* is a general category of cats, including cat types like *lion* and *tiger*, we can infer that:
>
> -   *Feline* is the hyponym of *lion* and *tiger*.
>
> -   *Lion* and *tiger* are the taxonyms of *feline*.

You can create true semantic word relationships by making keyword relationships at the word level using the following directional word equivalences:

| Direction | Word equivalence description |
| --- | --- |
| One-way synonyms | One-way synonyms relate words that involve a whole and part of the whole.<br><br>For example, a *finger* is a part of a *hand*. The word *hand* is the holonym of *finger* and *finger* is the meronym of *hand*. |
| Two-way synonyms | Two-way synonyms allow for bidirectional word equivalence for search terms.<br><br>For example, a search for *black pants* is equivalent to a search for *black trousers*. Both key phrases yield the same results although the order of the results might not be the same. |

With synonyms, you define replacements as an additional type of word-level relationship. This method omits any results matching the keyword and returns results for the replacement keyword. For example, potential bad results returned for keyword *chevy* are hidden, and are corrected with those for the replacement word *chevrolet*.

## Analyzers

Sitecore Search uses *analyzers* to ensure that searches return all relevant results instead of just exact matches. They are used to varying degrees in [filters](https://doc.sitecore.com/search/en/users/search-user-guide/refine-search-results.html "Refine search results"), personalization, suggestion blocks, [textual relevance](https://doc.sitecore.com/search/en/users/search-user-guide/display-relevant-results.html "Display relevant results"), and [sorting options](https://doc.sitecore.com/search/en/users/search-user-guide/refine-search-results/sorting-options.html "Sorting options").

Analyzers convert text input into a structured format that is optimized for search. They do this using the following three-step process:

1. If the analyzer uses character filters, it applies them. This means that certain characters are replaced or removed. For example, non-alphanumeric characters such as punctuation might be removed.
2. The analyzer *tokenizes* the search phrase. This means that the phrase is split up into smaller chunks called *tokens*. These tokens are usually single words, but can also be partial words or phrases.
3. If the analyzer uses token filters, it applies them. This means that the tokens are transformed using a variety of methods including applying synonyms, reducing tokens to their root words, removing stop words, and so on.

Sitecore Search offers the [pre-built analyzers](https://doc.sitecore.com/search/en/users/search-user-guide/optimize-keyword-searches/types-of-analyzers.html "Types of analyzers") listed in the following table:

| Basic analyzers | Advanced analyzers |
| --- | --- |
| Multi locale standard<br><br>Standard<br><br>Alphanumeric only<br><br>Keyword<br><br>Lowercase<br><br>Prefix match | Ngram based matching<br><br>Partial match<br><br>Shingle generator<br><br>Standard no stemmer<br><br>Exact prefix match |

> [!note]
> For more information about analyzers, see the [ElasticSearch documentation](https://www.elastic.co/guide/en/elasticsearch/reference/current/analysis.html) on the topic.