Eyeball Movement Detection Using Sector Line Distance Approach and Learning Vector Quantization

The shifting of education to learning supported by technology facing challenges. One of the challenges is to able to detect student learning processes, especially student affective state. Some studies implemented affective detection. However, the detection in obtrusive manner. Thus, there is a need to propose a method to detect student learning state in unobtrusive manner. This research propose a method to take advantage of eyeball movements as the measuring media as a feedback in the e-learning process. This study focuses in detecting eyeball movement. The eye movement to be detected in this study is in 5 directions center, upward, downward, leftward, and rightward. The eyeball movements is detected using the Sector Line Distance and classification method namely LVQ (Learning Vector Quantization). The result of the experiment shows that the average accuracy in detecting 5 eyeball direction shows a value of 81,72% accuracy. In addition, the proposed method also able to detect 5 directions of the eyeball movements including center. This result outperforming the previous method which only can detect 4 gaze of eyeball movements. Using center and 4 other directions of the eyeballs, the interesting value of the content can be measured and discovered.