Abstract
Test cases are an essential tool in software quality assurance: they ensure that code behaves as specified in the requirement. However, writing test cases does not have only benefits, it comes with a cost: the programmer has to formulate the test cases and maintain them when the tested source code changes. Particularly for start-ups or small enterprises such costs become prohibitive, which often prefer to invest their time into the development of new functionalities instead of testing. This paper explores the use of process-mining as an approach to create a model of how users interact with a system to a) generate test cases and b) prioritize them. Using process-mining, it is possible to mine from the user behaviour which parts of the system are the most used, in which order they are executed, generate test cases repeating user input, and prioritizing test cases.