Current speech recognition systems employ grammars to enable computers to
determine the meaning of user speech with reasonable accuracy and speed. The
JSpeech Grammar Format (JSGF) defines a platform-independent, vendor-independent
textual representation of one type of grammar, a rule grammar.