Horizontal Mouse Movements (HMMs) on Web Pages as Indicators of User Interest

Mouse events are widely used as implicit indicators of user attention on web pages. In this study, we investigated a particular pattern of mouse movements, Horizontal Mouse Movements (HMMs), consisting of series of mouse move events in the same horizontal direction, as indicators of users’ current interest. We formally defined HMMs and analyzed HMM activity on a sample website in English. We distinguished between LTR (Left to Right) HMMs and RTL (Right to Left) HMMs. LTR HMMs (in the reading direction of the sample website) were found to be more frequent than RTL HMMs (in the opposite direction). Then we investigated leaving web pages immediately after HMMs and found that they are much more frequent after an RTL HMM than after an LTR HMM. The difference can be explained by recent studies, which show that mouse movements in the reading direction are related to reading. Because reading indicates current interest in the web page content, the probability of leaving a web page immediately after LTR HMMs is lower. Accordingly, HMMs in the reading direction may serve as user interest indicators in educational technology, online learning, web analytics, and adaptive websites.