Web-Browser Access Through Voice Input and Page Interest Prediction

We propose a system that supports the interaction of handicapped people with web browsers for Internet navigation using voice. This system supports the interaction by offering a voice interface for implementing a pointing device. Moreover, it minimizes the number of voice commands by making predictions about the hyperlinks that the user is likely to select next. 1 The Web Browser Control Method The effective use of graphical interfaces is severely limited for most physically handicapped people, who find it very hard to control typical pointing devices. To provide a usable interface to people with such inabilities we conceived both a vocal device based on the simplest sounds that can be emitted clearly (vowels) and a system to aid user interaction by means of prediction of actions. The predictor aid is particularly useful since the voice input device has very limited control capabilities. The method we are currently developing controls one of the most popular web browsers, Netscape.