The Benefits of Agent-Oriented Goal Model in Developing Data Protection Application for Lecturers
Terje Mäesalu, Kerli Mooses · 2023
Ensuring that all teaching staff in the university follow data protection regulations is a great challenge. In developing a sociotechnical system that would make data protection topics easily accessible and help to avoid breaches, it is essential that the target groups are involved. One promising method to involve target groups who do not have an IT background is using agent-oriented goal models (AOM). The aim of the current study was to test the AOM usability in different stages of the requirements engineering when developing a data protection application for university lecturers. More specifically, we focused on how AOM supports the requirement elicitation with target groups, communication with funders, and the work of analysts during the requirement specification phase. Based on document analysis and pilot interviews, we created an initial goal model, which was validated with focus group interviews. The updated goal model was used to pitch the idea to funders. The feedback from focus group participants, funders and analyst concerning the AOM was analyzed. The results showed that AOM facilitated the interviews and activated the overall discussion. The presentation of AOM during the interviews helped to identify twice as many requirements than without using AOM. From a funder's perspective, the AOM made it easier to understand the necessity of the application. From an analyst perspective, AOM helped to keep focus during prioritization and creation of user stories. Overall, the AOM is a method that can be effectively implemented throughout the whole requirements engineering process.