Platform Independent Usage Mining Framework


Designing graphical user interfaces is a very important and difficult task. The main goal is to make them as intuitive and easy to use as possible. This is very important since it affects the users’ performance and their satisfaction with the application. It is an hard task that requires usability tests to be performed with end users of the application.

The objective of the system described in this document is to help developers and usability engineers assessing their user interfaces, not only during usability tests but also after the final application release. For this, an Android framework was developed, that can be implemented into any application from this platform, with the objective of logging user interactions with the application. After collecting usage data, it is sent to a server that automatically processes it using data mining techniques, extracting useful information that can be used to improve user interfaces. Besides usage data, the framework also collects visual information from the application, so it can be used to provide a visual analysis of the information extracted from usage data. Being able to see the extracted information visually, allows people that didn’t participate in the application’s development process to easily interpret its results.

To test this system, the framework was implemented in an Android application that was going to be used for some tests. This implementation was fairly easy, having arisen some minor problems that didn’t allow all the user interface components to be logged.
The application’s visual information was successfully retrieved. Since the application was used in a real case scenario, real usage data was collected and processed in order to extract information.
Without any usability expertise, it was possible to identify some user interface flaws based on the extracted information using this system.


Author: David Ribeiro

Type: MSc thesis

Partner: Faculdade de Engenharia da Universidade do Porto

Year: 2011