Designing Assistants for Prototyping Graphical User Interfaces (GUIs)
Early involvement of users into software development has proven to be a success factor for many software development projects. Especially, generating graphical user interfaces (GUI) prototypes has shown to be an effective method for user involvement. From the perspective of requirement elicitation and validation, GUI prototypes (e.g., as throw-away-prototypes) can serve as a catalyst for better understanding requirements with involved stakeholders. From the perspective of human-centered design, fast generation and testing of GUI prototypes, before more resource intense development steps are carried out, is key to not waste development resources.
Current tools and approaches in literature, but also from practice that support GUI prototyping are 1) scattered and not well integrated, 2) usually do not address novice users, who often perform GUI prototyping tasks in practice (e.g., when a backend software engineering also has to build a frontend GUI), and 3) while GUI prototyping is less resource intense in contrast to immediate software development, often GUI prototyping still involves a considerable amount of resources.
To address the presented problems, in this project we aim at 1) evaluating integrated adaptive systems that combine tailored approaches with tools from practice, 2) creating and evaluating approaches that support novice users through prototyping, testing of GUI prototypes and interpreting usability test data, and 3) proposing and evaluating LLM-based assistants that detect user story matching in GUI prototypes based on natural language, and recommend suitable components for non-implemented requirements.
