Hervé BLANCHON, Laurel FAIS
A Second Experiment on the
Understandability of Interactive
Disambiguation Dialogues
Abstract:We report on a second experiment that was carried out at ATR-ITL on the topic of
interactive disambiguation and more precisely, on the understandability of interactive
disambigua questions.
A first pilot experiment was carried out during the summer 1995 on the same topic
[Blanchon & Fais 1995]. This second experiment is the fruit of what we learned and the
results are far better.
Two classes of questions (human-like and machine-like) were proposed using two different
modalities (text and speech). The human-like questions were the easy ones and the
machine-like ones were the difficult ones. We had four groups of subjects, each subject
participating in one setting.
A text was read aloud by the subjects, who then answered 35 questions. The analysis of the
results shows that there is basically no significant difference among the four settings. This
tends to prove that the machine style questions, which could be produced automatically,
are not that hard to answer.