Automatic Support for Usability Evaluation

The main goal of this work is to propose a method to evaluate user interfaces using task models and logs generated from a user test of an application. The method can be incorporated into an automatic tool which gives the designer information useful to evaluate and improve the user interface. These results include an analysis of the tasks which have been accomplished, those which failed and those never tried, user errors and their type, time-related information, task patterns among the accomplished tasks, and the available tasks from the current state of the user session. This information is also useful to an evaluator checking whether the specified usability goals have been accomplished.