Profiling Exploratory Browsing Behaviour with a Semantic Data Browser

Semantic Web technologies are increasingly being adopted for aggregating Web data. Tools such as Semantic Data Browsers have been proposed to assist users to access and make sense of the vast semantic space. However, further investigations are needed to understand how users make use of the additional semantic features provided by these new breed of browsers and their effectiveness in supporting exploration of a domain. Measurements of browsing behaviour in a semantic space are also needed. Using the log data from a semantic browser (MusicPinta) for the music domain, this paper takes the first step in profiling browsing behaviour of users in a semantic space and compares the outcome against their task performance. Two exploratory search tasks were designed for the experiment. Movements in terms of users traversing the provided semantic links in the browser were captured and the patterns of clicks between abstract and concrete concepts were analysed.