Masterarbeit
From UI Tests to Complexity Insights: Predicting Interaction Load in Web Applications
Completion
2025/12
Research Area
Students
Bukuroshe Xhamxhiu
Advisers
Dr.-Ing. Sebastian Heil
Description
Automatic evaluation of the user interface (UI) can provide valuable insights on how usable and attractive a web application is for users, directly affecting the potential success of it, decreasing the effort for time-consuming, expensive and massive user studies. Interaction complexity and task load are two important factors that influence the usability of a web application. UI Tests are used to ensure that the application behaves as expected and therefore are representations of the user interfactions with the web user interfaces. Therefore, they represent a potential source of information for assessing the complexity of user interfaction of web applications.
This thesis aims exploring the potential of using UI tests for complexity and task load prediction. Previous work has shown that limited availability of public code repositories with a high coverage and consistency of UI tests are a major obstacle for systematic experimentation. Therefore, the approach tested in this thesis will apply methods of automated generation of UI tests to create the basic input for the assessment of complexity and task load. To test the feasibility of this approach, the complexity and task load results need to be systematically compared to those from traditional manual assessment in a user study.
The objective of this thesis is to create a solution or the combination of existing approaches to solve the problem of automated complexity/task load assessment of web user interfaces through analysis of automatically generated UI tests as described above. This includes the analysis of the state of the art of existing methods of UI complexity and load measurement and UI test generation as well as the demonstration of the solution by prototypical implementation and a suitable comparative evaluation with a user study.
