SubsetMatchingMode Enumeration

Enumerates values of subset matching mode.

Namespace:  Microsoft.Speech.Recognition
Assembly:  Microsoft.Speech (in Microsoft.Speech.dll)

Syntax

'Declaration
Public Enumeration SubsetMatchingMode
'Usage
Dim instance As SubsetMatchingMode

public enum SubsetMatchingMode

Members

Remarks

Using the Microsoft Speech Platform SDK 11, you can construct a grammar that successfully recognizes a phrase even though only a subset of its contents is found in the speech input.

By default, a recognition engine requires an exact match against an entire phrase. The recognition engine can also match partial phrases according to parameters defined by a SubsetMatchingMode member used as an argument to the GrammarBuilder(String, SubsetMatchingMode) constructor or the Append method.

The following illustrates how each of the SubsetMatchingMode members affects recognition when used in a grammar that contains the phrase "a car the truck a boat that plane".

OrderedSubset

This mode indicates that a subset of the phrase will be used to successfully recognize the entire phrase if the following are true:

• One or more words in the grammar phrase are recognized in the speech input.

• The relative order of those recognized words is the same as in the phrase.

• Speech input may consist of only prepositions and articles.

Adding the phrase "a car the truck a boat that plane" to a grammar using OrderedSubset mode produces the following result on recognition:

OrderedSubsetContentRequired

This mode indicates that a subset of the phrase will be used to successfully recognize the entire phrase if the following are true:

• One or more words in the grammar phrase are recognized in the speech input.

• The relative order of those recognized words is the same as in the grammar phrase.

• Matched speech input cannot consist of only prepositions and articles.

Adding the phrase "a car the truck a boat that plane" to a grammar using OrderedSubsetContentRequired mode produces the following result on recognition:

Subsequence:

This mode indicates that a subset of the phrase will be used to successfully recognize the entire phrase if the following are true:

• One or more words in the speech input occur in the exact same sequence as in the grammar phrase.

• Matched speech input may consist of only prepositions and articles.

Adding the phrase "a car the truck a boat that plane" to a grammar using Subsequence mode produces the following result on recognition:

SubsequenceContentRequired

This mode indicates that a subset of the phrase will be used to successfully recognize the entire phrase if the following are true:

• One or more words in the speech input occur in the exact same sequence as in the grammar phrase.

• Matched speech input cannot consist of only prepositions and articles.

Adding the phrase "a car the truck a boat that plane" to a grammar using SubsequenceContentRequired mode produces the following result on recognition:

Examples

Below is a test routine which generates four Grammar objects, using the same phrase, for each of the SubsetMatchingModes. These Grammar objects can then be tested.

private void CreateSubsetMatchTest() 
{
  foreach ( Microsoft.Speech.Recognition.SubsetMatchingMode mode 
              in Enum.GetValues(typeof(Microsoft.Speech.Recognition.SubsetMatchingMode)))
  {
    GrammarBuilder gb = new GrammarBuilder("a car the truck a boat that plane",mode);
    Grammar grammar = new Grammar(gb);
    grammar.Name = mode.ToString();
    grammar.Enabled=true;
    _recognizer.LoadGrammar(grammar);
  }
}

See Also

Reference

Microsoft.Speech.Recognition Namespace

GrammarBuilder