Improving spatial data usability by capturing user interactions

Abstract

In this paper we present a flexible architecture for both desktop and mobile systems designed to measure users' implicit levels of interest in spatial information, based on their mouse movements. Users interact with spatial objects through a GIS interface. Our system monitors their actions and compares the spatial location of their mouse interactions during a session to the underlying spatial information displayed to the user. By making this comparison we implicitly determine a user's interests and disinterests. This information can be used to build up a user profile on a session by session basis, allowing the system to personalise future datasets to present to the user based on his profile, reducing information overload, and increasing the ratio of user-relevant content. The case study for our system concerns archaeological information for burial tombs in the ancient city of Tarquinia.