Contextualising user interfaces for complex systems

This thesis addresses the difficulties that people have in interacting with complex, computer-based systems. The fields of intelligent and adaptive interfaces and agent-based systems are surveyed and critiqued to identify how intelligent human-computer interfaces can be used to improve interaction. The need to manage context is identified as a key element in intelligent interfaces used to manage complexity. A model of contextualization is developed to encompass a range of interface design and implementation paradigms, with the objective of improving the design of dynamic interactive systems. Viewing the process of contextualization as part of the interaction process provides a powerful conceptual methodology for the design of agent-based intelligent user interfaces. A model of contextualization is developed consisting of several components that are intended to promote contextualization in user and interface. An experimental evaluation of these components shows that elements of dialogue instigation and adaptation of the user interface via user preferences provide the expected performance advantages in both objective and subjective evaluations. In addition, the experiments show that contextualization is affected by factors such as the personality of the user. The interaction of the various components of the model of contextualization is discussed and proposals for future work are presented.

Name dc:type.qualificationname

d_ph

Level dc:type.qualificationlevel

d_ph

Grantor dc:publisher.institution

University of Birmingham

Year dc:date.issued

2000

Author dc:creator

• Meech, John Foster

• QA75 Electronic computers. Computer science
• BF Psychology