| Abstract |
Over the past decade, Graphical User Interfaces (GUIs) have evolved and largely replaced text-based interfaces in the operating systems of personal computers. However, a sizable part of the user population experiences significant difficulties in accessing GUIs. This thesis addresses the problems faced by motor-impaired users in accessing the traditional input devices that are used to control GUIs, such as the keyboard and pointing devices (e.g. mouse). Existing approaches to facilitate accessibility involve hardware and/or software adaptations that deploy specialized input devices, such as binary state switches and involve scanning techniques. Scanning techniques are based on the principle of selecting interaction options and objects that are being highlighted sequentially. The user selects an option at each dialogue step by effecting a selection switch. An indicator ``scans`` the possible options in a predefined order. The user can activate the highlighted option at a predefined time interval. The movement of the indicator can be controlled by the user - through a second binary switch - or may be automatically activated after a predefined time delay. Apart from their use in systems that support people with motor impairment, scanning techniques can be used in any interaction system where devices need to be very simple and can not direct selection. This thesis presents a new approach to offering access to graphical applications to people with motor disabilities. Following this approach, access is provided through a number of internal functional expansions that are proposed for the common interaction objects of the most widespread Graphical User Interface system (Microsoft Windows). New dialogue methods have been designed for each of the common interaction objects. The methods are based on scanning techniques and can be employed through the use of one or two binary switches, according to the user's capabilities and preferences. The main advantage of this approach is that interaction can thus become more flexible and straightforward, since the dialogue is specifically designed for each interaction object. In that way, many of the problems that existing approaches suffer from can be eliminated. Such problems are the need for the user to focus his/her attention to more than one points simultaneously and the presence of a virtual (on-screen) keyboard that uses of a large portion of the screen space (real estate) while no text input is being performed. The above dialogue methods have been embedded into a set of programming classes that correspond to interaction objects. These are derived from the interaction objects' base classes provided by the scanning dialogue method is selected and controlled. Those parameter variables are common for each interaction object. User interface developers can use the new augmented classes similarly to the way they used the base classes and then be activated, while the corresponding object parameters. The scanning dialogue can then be activated, while the application can still be accessible through the classic devices. Therefore applications that use the new base classes for their interaction objects can be accessed both by the classic input devices and by switches. This follows the principle of ``designing for all users``. The scanning dialogue is now internally supported by the application without the need of external adaptations. The design and development of the application is only carried out once, and alternative dialogue methods are already embedded. The new classes have been employed in the development of a web browsing application, the "AVANTI Browser", which was one of the outcomes of the ACTS AC042 AVANTI project (Adaptive and adaptable Interactions to Multimedia Telecommunications Applications), partially funded by the European Commission. This application was used for the evaluation of the new methods of the embedded scanning dialogue, the objective being to identify efficiency of the dialogue methods, as well as the strong and weak points of the approach. The evaluation produced satisfactory results. This report reviews the current state of the art, providing the motivation of the work and the rationale of the proposed approach. An analysis of the new dialogue methods for interaction objects in graphical User Interfaces follows and the development and use of the new programming classes are described. The evaluation process and its results are described and conclusions along with recommendations for future work are provided.
|
Τεχνικές Σάρωσης για Αντικείμενα σε Γραφικά Περιβάλλοντα Αλληλεπίδρασης
