| Popis: |
German universities are facing an intense, competitive environment caused by globalization, digitalization, and public sector reforms. The latter also gave the universities more decision-making autonomy, which goes hand in hand with more responsibilities, but also with the possibility of individualizing their strategy. This thesis examines how German universities can use Data Mining techniques to extract useful information from their available data resources to address these current challenges by supporting management decisions. The use of Data Mining methods in education is called Educational Data Mining. Research in this area has so far focused mainly on supporting students and lecturers. This thesis focuses on researching the benefits of Educational Data Mining for university management, which has been mentioned several times in various Educational Data Mining studies but has not been studied in detail so far. After discussing the most important challenges faced by German universities, their current tasks and objectives were examined. A framework model was then developed that illustrates how the results of two specific Data Mining projects can help universities tackle the challenges and accomplish their tasks. The selected Data Mining projects are dropout analysis and enrollment prediction because the student and applicant data are available to all the German universities. The proposed framework model was verified with two case studies in which the specified analyses were carried out at a German university of applied sciences. To build well-performing models, several Data Mining methods were used and compared. Subsequently, the results were discussed with representatives from the case university, and suggestions were made how the information generated could be included in the decisions of the university administration. It has been shown that German universities can use their data resources to support their management activities. An overview of this support was presented in the form of a framework model that is not only a first attempt to close the existing research gap in the field of EDM but should also mo-tivate university decision-makers to use their existing data resources. Therefore, the presented thesis can stimulate further research that combines the results of EDM projects with managerial decisions to increase the efficiency of educational institutions. In addition, university administrators can be inspired to use all available resources to ensure their long-term success. |
