Speech recognition software "enables data to be input by voice." This kind of software can assist people who have difficulty navigating computers with keyboards and mice. Speech recognition software can also be useful for people whose oral skills are more advanced than their writing/typing skills.
Some speech recognition software allows users to totally control their computers with their voice, including performing commands like opening applications, creating text, sending emails, browsing the web, etc. Some speech recognition software can create text from voice but requires users to also use the computer mouse and keyboard for extra functionality.
Information from "Speech Recognition Software" in The Hutchinson unabridged encyclopedia with atlas and weather guide.
Speech recognition software came to prominence in the 1970s. Since the 1970s considerable progress has been made in the effectiveness of this type of software. Before 1990, the vocabulary of this software was usually around 1,000 words. Today the vocabulary of speech recognition software can be above 65,000 words. Over the years, research in this area has turned away from how to command certain tasks to be performed on a computer to making computers better understand continuous speech.
Information from Speech recognition algorithms using weighted finite-state transducers by Takaaki Hori and Atsushi Nakamura