Nestor Garay Vitoria 2000

Finished

Description:

The main objective of this work is studying word prediction or anticipation methods to apply them to the user interface design in Augmentative and Alternative Communication systems, in order to enhance the communication speed of people with disabilities. First, a wide bibliographic revision is made to determine the background, before an exhaustive study of the factors that have influence in the linguistic prediction in general. In this way, the characteristics that may be used to get best results are established. Likewise the influence of the user interface in the efficiency of predictors is studied in conjunction with the qualities to enhance their usability. Moreover, the parameters that serve to measure the quality of the anticipators and the methods to evaluate them are analysed. Then, and taking into consideration the language dependency that anticipatory methods have, some separate studies for languages with low and high flexion degree have been made. The designed methods are tested for the Spanish and the Basque languages, respectively. These methods are formally specified, to allow a good understanding and their adaptation to different environments. Finally, their integration in some human-computer interfaces is shown. These devices are currently under evaluation by people with disabilities.

Participants:

Nestor Garay Vitoria

Author:

Nestor Garay Vitoria

University:

Universidad del País Vasco-Euskal Herriko Unibertsitatea

Center:

Facultad de Informática

Rating:

Sobresaliente Cum Laude. Premio Extraordinario de Doctorado de la UPV-EHU en el 2002

Reading year:

2000