Here be apples and oranges…. Generally speaking, there are two different components. One is the speech recognition component, and the other is the natural language understanding component. The speech recognition component converts speech into text, and feeds text into the natural language understanding component.
. . .
Wikipedia lists a variety of, such as the ; however, to my knowledge, spoken language corpora as such are generally not applied to AI. (There is evidence of older research that tried going in that direction.)
There is a field of study known as(see also ), which encompasses , , and .