A method of human-computer interaction in which computers listen and interpret human speech to produce written or spoken outputs.