Search functions should be used when performing full-text search, namely when the `MATCH` or `QUERY` predicates are being used. Outside a, so-called, search context, these functions will return default values such as `0` or `NULL`.
Elasticsearch SQL optimizes all queries executed against {{es}} depending on the scoring needs. Using [`track_scores`](/reference/elasticsearch/rest-apis/sort-search-results.md#_track_scores) on the search request or [`_doc` sorting](/reference/elasticsearch/rest-apis/sort-search-results.md) that disables scores calculation, Elasticsearch SQL instructs {{es}} not to compute scores when these are not needed. For example, every time a `SCORE()` function is encountered in the SQL query, the scores are computed.
**Description**: A full-text search option, in the form of a predicate, available in Elasticsearch SQL that gives the user control over powerful [match](/reference/query-languages/query-dsl/query-dsl-match-query.md) and [multi_match](/reference/query-languages/query-dsl/query-dsl-multi-match-query.md) {{es}} queries.
The first parameter is the field or fields to match against. In case it receives one value only, Elasticsearch SQL will use a `match` query to perform the search:
```sql
SELECT author, name FROM library WHERE MATCH(author, 'frank');
author | name
---------------+-------------------
Frank Herbert |Dune
Frank Herbert |Dune Messiah
Frank Herbert |Children of Dune
Frank Herbert |God Emperor of Dune
```
However, it can also receive a list of fields and their corresponding optional `boost` value. In this case, Elasticsearch SQL will use a `multi_match` query to match the documents:
```sql
SELECT author, name, SCORE() FROM library WHERE MATCH('author^2,name^5', 'frank dune');
The `multi_match` query in {{es}} has the option of [per-field boosting](/reference/query-languages/query-dsl/query-dsl-multi-match-query.md) that gives preferential weight (in terms of scoring) to fields being searched in, using the `^` character. In the example above, the `name` field has a greater weight in the final score than the `author` field when searching for `frank dune` text in both of them.
Both options above can be used in combination with the optional third parameter of the `MATCH()` predicate, where one can specify additional configuration parameters (separated by semicolon `;`) for either `match` or `multi_match` queries. For example:
```sql
SELECT author, name, SCORE() FROM library WHERE MATCH(name, 'to the star', 'operator=OR;fuzziness=AUTO:1,5;minimum_should_match=1')
**Description**: Just like `MATCH`, `QUERY` is a full-text search predicate that gives the user control over the [query_string](/reference/query-languages/query-dsl/query-dsl-query-string-query.md) query in {{es}}.
The first parameter is basically the input that will be passed as is to the `query_string` query, which means that anything that `query_string` accepts in its `query` field can be used here as well:
```sql
SELECT author, name, SCORE() FROM library WHERE QUERY('name:dune');
The query above uses the `_exists_` query to select documents that have values in the `author` field, a range query for `page_count` and regex and fuzziness queries for the `name` field.
If one needs to customize various configuration options that `query_string` exposes, this can be done using the second *optional* parameter. Multiple settings can be specified separated by a semicolon `;`:
```sql
SELECT author, name, SCORE() FROM library WHERE QUERY('dune god', 'default_operator=and;default_field=name');
**Description**: Returns the [relevance](https://www.elastic.co/guide/en/elasticsearch/guide/2.x/relevance-intro.html) of a given input to the executed query. The higher score, the more relevant the data.
When doing multiple text queries in the `WHERE` clause then, their scores will be combined using the same rules as {{es}}'s [bool query](/reference/query-languages/query-dsl/query-dsl-bool-query.md).
