swirec_acoustic_adapt_root

Specifies the storage location of self-learning files. Controls sharing of models across the server, tenants, or applications.

Value

A directory path on the local machine

DEFAULT: (empty)

By default, this parameter is not set:

  • If companies are configured on the host, the recognizer adapts separate models for each company’s applications.
  • If no company is configured, the recognizer adapts models for the recognition server.

How to set

In the application's session xml in the <speechrecog> section of <sessionparameters>.

Usage

Occasionally used for systems that host applications for more than one tenant and application. By defining the location of self-learning files, you control the scope of learning across tenants and applications.

Controls per-tenant adaptation. Tenant-adapted acoustic models are stored in the tenant data directory.

The recognizer stores all acoustic data beneath this location. This includes acoustic models and adaptation statistics files.

Tenant-based adaptation—By default, the system adapt models for all of a tenant’s applications.The definition in a session.xml looks like this:

<param name="swirec_acoustic_adapt_root">
  <value> ${APP_DATA_DIR}/acoustic_adapt </value>
</param>

Server-based adaptation—To adapt models for all applications on a single Nuance recognition server, use $SWISRSDK as the root in the path. The definition in a session.xml looks like this:

<param name="swirec_acoustic_adapt_root">
  <value> ${SWISRSDK}/acoustic_adapt </value>
</param>

Seldom needed. It’s also possible to adapt models for individual applications running on a single recognition server:

<param name="swirec_acoustic_adapt_root">
  <value> ${SWISRSDK}/MyApp2/acoustic_adapt </value>
</param>

For an overview of adaptation, see Controlling the self-learning recognizer.

Related parameters: