Interface design is often constrained by the limited scope and resource-intensive nature of conventional user studies. We aim to unburden this process by introducing an automated user modelling framework that continuously injects design guidelines into the development process. We present a pipeline that converts a given user interface design into a widgetised data structure, executes a performance simulation based on the cognitive model of a user, and analyses its output to give design guidelines. We introduce the research methodology employed to create the model, implementation details of the model, and initial results from its validation. These include the dynamics of age-based modelling, the temporal integrity of the output of the cognitive model, and indications of the accuracy of the overall design guidelines produced.

Digital devices are often restricted by the complexity of their user interface (UI) design. While accessibility guidelines exist that reduce the barriers to access information and communications technology (ICT), guidelines alone do not guarantee a fully inclusive design. In the past, iterative design processes using representative user groups to test prototypes were the standard methods for increasing the inclusivity of a given design, but cognitive modelling (the modelling of human behaviour, in this instance when interacting with a device) has recently become a feasible alternative to rigorous user testing (John and Suzuki 2009). Nonetheless, many models are limited to an output that communicates little more than the assumed time the modelled user would require to complete the task given a specific way of doing so (John 2011). This chapter introduces a novel approach that makes use of the overlay of user modelling output (timings) onto a graphical representation of an entire UI, thereby enabling the computation of new metrics that indicate the relative inclusiveness of individual screens of the UI.

Touchscreen devices are often limited by the complexity of their user interface design. In the past, iterative design processes using representative user groups to test prototypes were the standard method for increasing the inclusivity of a given design, but cognitive modeling has potential to be an alternative to rigorous user testing. However, these modeling approaches currently have many limitations, some of which are based on the assumptions made in translating a User Interface (UI) into a definition file that cognitive modeling frameworks can process. This paper discusses these issues and postulates potential approaches to improvements to the translation procedure.