Meaningful Object Extraction for Booking Website

This paper attempts to propose a novel approach that can obtain the interested data from extracting the web objects. Through the object extraction and analyzing the gaze distributions, the meaningful object on webpage can be acquired in corresponding to the viewer’s real gaze behavior. In order to locate the meaningful object on webpage, the proposed approach includes two parts: the interested object extraction and the mapping model of the meaningful object content. In addition, the proposed method can overcome the content uncertainty caused by the mouse events when the audience scrolls or clicks on the mouse. The experimental results show that the proposed approach returns a good performance.