Dalarna University aspires to improve the student admittance process as it strives to ensure efficient use of public funds. The aim of Dalarna University in the student admittance process is to admit candidates that are likely to be successful in their studies. There have been two selection procedures for the Business Intelligence and Data Science programmes: the default procedure, which selects from all eligible candidates randomly, and the motivation letter procedure in which the university bases their selection on a motivation letter that each eligible candidate must provide. it is of interest whether the motivation letter procedure outperforms the default procedure as a selection instrument. For that, it is needed to analyse student data. Student data is stored in foremost two national databases: NyaWebben and Ladok. These databases are not conducive for analysis. The main objective of this researchis to create a data template that can store data from these systems and can also be used to analyse student data in order to evaluate the performances of the selection instruments. To demonstrate the usability of the data template, we formulated two demonstration questions that relate to the performance of the selection instruments that we tentatively tried to answer: 1) Do the motivation letter intakes show a higher degree completion rate and strength of degree completion than the default intakes? 2) How do the assigned motivation letter scores relate to degree completion rate and strength of degree completion? Firstly, we identified which variables were to be extracted from NyaWebben and Ladok and established what KPIs would serve as the working indicators used to analyse the performances of the selection instruments. From a literature review, we deduced that degree completion and strength of degree completion were the KPIs that considered the overarching perspectives of universities, students, the labourmarket, and the government on what constitutes a favourable educational outcome. After that, we employed a manual extraction method to store the data in a spreadsheet; our data template of choice. We then build multiple interfaces and automation procedures for increased efficiency and ease of use. The data template was made analysis-ready through data cleaning, categorisation, and formatting of the data. We furthermore conducted a usability test to assess the ease of use of our data template and the results were positive. We learned that there are still too few data points because of an insufficiently long follow-up time, even when considering only first semester completion, to conclusively answer the demonstration questions and that our results were not statistically significant. The demonstration questions can therefore only be answered in time once more student intakes are added to the data template. However, we found that the implications of the data template are manyfold, since we considered a wide range of variables, including those outside the scope of the demonstration questions, that are ready to be analysed via our data template.

As the popularity of the master’s in data science at Dalarna University increases, so does the number of applicants. The aim of this thesis was to explore different approaches to provide an automated motivation letter scoring system which could emulate the human judgement and automate the process of candidate selection. Several steps such as image processing and text processing were required to enable the authors to retrieve numerous features which could lead to the identification of the factors graded by the program managers. Grammatical based features and Advanced textual features were extracted from the motivation letters followed by the application of Topic Modelling methods to extract the probability of each topics occurring within a motivation letter. Furthermore, correlation analysis was applied to quantify the association between the features and the different factors graded by the program managers, followed by Ordinal Logistic Regression and Random Forest to build models with the most impactful variables. Finally, Naïve Bayes Algorithm, Random Forest and Support Vector Machine were used, first for classification and then for prediction purposes. These results were not promising as the factors were not accurately identified. Nevertheless, the authors suspected that the factors may be strongly related to the highlight of specific topics within a motivation letter which can lead to further research.

Manually reading, evaluating, and scoring motivation letters as part of the admissions process is a time-consuming and tedious task for Dalarna University's program managers. An automated scoring system would provide them with relief as well as the ability to make much faster decisions when selecting applicants for admission. The aim of this thesis was to analyse current human judgment and attempt to emulate it using machine learning techniques. We used various topic modelling methods, such as Latent Dirichlet Allocation and Non-Negative Matrix Factorization, to find the most interpretable topics, build a bridge between topics and human-defined factors, and finally evaluate model performance by predicting scoring values and finding accuracy using logistic regression, discriminant analysis, and other classification algorithms. Despite the fact that we were able to discover the meaning of almost all human factors on our own, the topic models' accuracy in predicting overall score was unexpectedly low. Setting a threshold on overall score to select applicants for admission yielded a good overall accuracy result, but did not yield a good consistent precision or recall score. During our investigation, we attempted to determine the possible causes of these unexpected results and discovered that not only is topic modelling limitation to blame, but human bias also plays a role.