Abstract: In the psychology and cognitive science, intentional modeling and recognition are considered
to be the new models for creating human-computer interaction. Explaining user’s implicit intention
recognition has been a hotspot in interaction design research. In this study, an implicit intention
recognition system based on eye movement pattern and pupil size change was proposed. A baseline
model was established to measure pupil size change. Eye movement experiment was conducted with a
juicer as an example. Finally, the implicit intention is divided into two types: unintentional intention and
purposeful intention, by using the classifier to process the experimental data. The experimental results
show that the discussion of this paper has some reference significance to the research of implicit
intention recognition, and provides support for product or system research and development.

Key words: human-computer interaction, eye movement pattern, implicit intention, classifier