Wizard of Oz experiment

from Wikipedia, the free encyclopedia

In human-computer interaction, a Wizard of Oz experiment is an experiment in which a person ( test subject ) assumes that he is communicating with an autonomous (in the sense of artificial intelligence ) system. In reality, however, another person is hidden who is generating the system's reactions.

Such experiments are conducted to gather possible responses from potential users of a system that is just being developed. This can be, for example, a dialog system that is to be used in a domain for which no experience data is yet available.