... information was logged, e.g. the transcriptions of the spoken utterances, the wizard sdatabase query and the number of results, the screenoption chosen by the wizard, and a rich set of con-textual dialogue ... bootstrapped from4 The ratings are normalised as some of the questions wereon different scales.a small amount of Wizard- of- Oz data, and we evalu-ated the result with real users. The use of WOZ dataallows ... side, the noise model de-fines the likelihood of the user accepting or rejecting the system’s hypothesis (for example when the sys-tem utters a confirmation), i.e. in 30% of the cases the user...