Interaction et contexte dans les interfaces zoomables

Many interactive computer systems use menus as an important part of their interface. Menus allow users to select operations but not to control their execution. A second interactor, such as a dialog box, has to be used to control the chosen operation and thus complete the interaction. This decomposition of what is a single action from the user’s point of view into two distinct steps slows down interaction with computer systems. This thesis proposes a new contextual pop-up menu, called a Control Menu, that includes proportional control of the chosen operation with immediate feedback. Using this menu gives a more fluid control of complex interfaces and has the advantage of an expert form of use that is very similar to, and thus easily learnt from, the novice usage. Interaction with databases and navigation within large information spaces are important tasks in many applications. Many visualization systems cause user disorientation as users find it difficult to understand their position within the information space and to locate desired information. This thesis proposes several new contextual aids for Zoomable User Interfaces that address these issues. The first aid, a hierarchical view of the information space, helps users understand their current position and the location of the desired information, and accelerates navigation. The second type of aid uses dynamically generated transparent and temporary views that are created and controlled by users in a single gesture. These interactive views overlay the current view of the focus with contextual or historical information which shows users what surrounds the current view or the route taken to arrive at that view. The effective use of these new aids requires a tight coupling between interaction and presentation which is achieved via the use of Control Menus.