the recognition of human speech by means of a computer-based process. Recognition in this context is assessed via the written representation of the words spoken. Unlike automated language understanding, ASR does not require the program to understand the meanings of the words recognized.