A Comparative Study Of Web - Based Visual Search And Eyeless Browsing

Internet significantly affects the way people seeking for information. As thehuman-computer interaction interface, web page gathers a large number of information andprovides people with a convenient platform of conveying information. And users acquireinformation on the web by visual behavior. A good man-machine interface will help to improvethe user’s Internet experience and enhance the efficiency of information retrieval, so it isnecessary to deeply research on the visual behavior characteristics of users’ viewing the webpage.There are two kinds of visual behavior when users view the web page—Web page visualsearching and browsing without a purpose. This paper analyzes and organizes previousliteratures related to the two kinds of visual behavior. And based on this, we use eye trackingtechnology to study the eye movement characteristics of Web page visual searching andaimlessly browsing. By analyzes the difference of eye movement between these two kinds ofvisual behavior, find the common features and differences of these two kinds of visual behavior.In experiment, this paper selects63participants and assigns participants to two groupsrandomly (visual search31, aimlessly browse32). This research takes2(web page types:animation and no animation)×2(task type: visual searching and aimlessly browse)×9(ninetargets’location: nine area in a web page) three factors mixed design By analyzing the dependentvariables: the response time of the participants in visual search and the parameters of eyemovement (fixation point number, fixation duration and average pupil diameter),discusses howthe two types of web pages (animation and no animation) influence participants’ visual searchefficiency and the similarities and difference of eye movement characteristic parameters in webvisual search and aimlessly browse.Research conclusions can be summarized as follow：(1) When search targets are distributed in different areas of webpage, the search efficiencyand eye movement parameters of the subjects both show significant differences. When the targetis in the upper-left corner or the upper part of the page, the efficiency of visual searching ishighest. However, when the target is in the upper-right corner or bottom-right corner， theefficiency is quite low. The center, the left, and the upper-left of web page is the user of “visualpreference area”, and the bottom-right and the upper-right of web page is the user of “visualblind area”. The sequence of web browsing preferences is: center, upper-left, left, middle,bottom-left, right, bottom-right, upper-right.(2)When visual searching web page, subjects first look at the middle part of webpage, thenquickly saccades around the peripheral webpage, taking visual search strategy which pay more attention to peripheral area. When aimlessly browsing web pages, subjects look at webpage byrandomly browsing, the distribution of first fixation point is more dispersed. Eye movementindexes such as fixation point number, fixation duration and the average pupil size in web pagevisual searching are more than aimlessly browsing.(3)Different type (animation and no animation) of web page has influence to visualsearching. Participants in the animation web page search efficiency were significantly lower thanthe web page without animation. However, aimlessly browsing, the fixation point number of twotype of web page has no difference. When visual searching web page, subjects pay little attentionto graphic animation; However, when aimlessly browsing, subjects pay more attention to graphicanimation.(4)In the visual searching research, most of the subjects have no obvious impression ongraphic animation; However, in aimlessly browsing research, all of the subjects have a obviousmemory on animation image, but the memory is not accurate.