[DOCS] Option to disable device type extraction in user agent process…

…or (#71626) (#77705)

docs/reference/ingest/processors/user-agent.asciidoc

@@ -21,6 +21,7 @@ The ingest-user-agent module ships by default with the regexes.yaml made availab
 | `target_field`         | no        | user_agent                                                                                      | The field that will be filled with the user agent details.
 | `regex_file`           | no        | -                                                                                               | The name of the file in the `config/ingest-user-agent` directory containing the regular expressions for parsing the user agent string. Both the directory and the file have to be created before starting Elasticsearch. If not specified, ingest-user-agent will use the regexes.yaml from uap-core it ships with (see below).
 | `properties`           | no        | [`name`, `major`, `minor`, `patch`, `build`, `os`, `os_name`, `os_major`, `os_minor`, `device`] | Controls what properties are added to `target_field`.
+| `extract_device_type`  | no        | `false`                                                                                         | beta:[] Extracts device type from the user agent string on a best-effort basis.
 | `ignore_missing`       | no        | `false`                                                                                         | If `true` and `field` does not exist, the processor quietly exits without modifying the document
 | `ecs`                  | no        | `true`                                                                                         | Whether to return the output in Elastic Common Schema format. NOTE: This setting is deprecated and will be removed in a future version.
 |======