Appendix A: Hark Grammar

The grammar used in the voice recognition system is described below in standard Backus-Naur notation. Numbers in the range from 1 to 99,999 were recognized by this grammar. In addition, simple variations on the normal pronunciation were allowed, such as "forty-two thousand and one hundred" for "forty-two thousand, one hundred." Encompassing these variations improved flexibility and thus improved the accuracy of the recognition system.

NUMBER => [ (TENNUMBER | DIGIT) thousand] and [DIGIT hundred] and [TENNUMBER]



TENNUMBER => TENS | DIGIT

TENS => TENSXTY DIGIT | TENSXTY | TEENS

TENSXTY => twenty | thirty | forty | fifty | sixty | seventy | eighty | ninety

TEENS => ten | eleven | twelve | thirteen | fourteen | fifteen | sixteen | seventeen | eighteen | nineteen

DIGIT => one | two | three | four | five | six | seven | eight | nine



On to next section: Appendix B