Due to the limited availability of human academic advisors, and the high demand for academic advising by students, students’ needs are not satisfied. In a digital world, collecting data has become increasingly important. Algorithms can be used for analysing data and building predictive models. More and more industries are using recommender systems to improve their services and personalize recommendations to satisfy every customer’s need better. Compared to humans, algorithms can also consider implicit data, which refers to information that is not explicitly stated but can be deducted from available data.

Researchers have introduced different approaches for building course recommender systems to support the academic advising process. The focus was on recommending courses which led to the highest student performance. Therefore, the goal of this master’s Thesis is to analyse course enrolment data from the University of applied sciences Northwestern Switzerland to find courses and course combinations which led to negative performances.

Furthermore, the discovered negative course-taking patterns are used to build a course recommender system prototype to support students in their enrolment process and discover insights for the curriculum planning process.