About the language model

About

The language model is a collection of statistical information associated with a specific vocabulary that helps predict the words that are most likely to occur in a user's speech.

Dragon uses the language model, acoustic model, and your vocabulary to achieve high recognition accuracy.

When it's updated

Dragon updates your language model when you run the following:

  • Accuracy Tuning
  • Learn from specific documents

For more information, see About Accuracy Tuning and Learn from specific documents.

 

Related procedures

Additional information

The Acoustic and Language Model Optimizer dialog boxAbout Accuracy Tuning
The Acoustic and Language Model Optimizer Scheduler About acoustic models
 About improving accuracy