The Human Informational Intent Classification by Eye Movement Analysis

Objective: The aim of this study is to figure out the presence or absence of human intent and to classify informational intent by analyzing eye movement information. Background: A Human being usually has intent that is an idea or plan to do something. In this topic, research for web query intent is dominated. But recently, intent study is proceeded in various fields like driver intent, detecting intent etc. Method: Experiment 1 verifies that intent is present or absent by using eye tracker and intent tasks that arousing intent or not. In Experiment 2, informational intent is divided into cognitive intent and affective intent. And the eye movement patterns are compared in each condition by using eye tracker and tasks that arousing intent. Results: The result of experiment 1 shows significant differences in average fixation duration, average pupil size under the condition of presence of human intent. And experiment 2 reveals significant differences between cognitive intent and affective intent in average fixation count, average fixation duration, average pupil size. Also, the time to complete cognitive and affective intent tasks shows significant differences. And the front part of entire stimulus presentation time proves significant differences between cognitive and affective intent condition in average fixation count and duration, but the later part doesn’t show significant differences. Conclusion: This paper can offer differences in eye movement patterns under human intent condition. And it is useful to verify human intent. Application: This study demonstrates the method to figure out human intent by using eye movement patterns, so that it is helpful to use more conveniently and correctly a computer or machine in HCI and HMI environment.