Theses procedure

Assignment

After the first discussion with the supervisor on the topic of the thesis has taken place, but before the student works on the actual topic, an assignment is prepared by her.

The goal of the assignment is to demonstrate that the research questions and core tasks of the thesis have been understood. It should illuminate the scientific context of the thesis, what is the scope of the thesis and what scientific methods will be used. An assignment is one to two pages in length and covers the following points:

Motivation: Why is the research question relevant? What is the challenge? Where could a potential solution be applied?
Embedding in existing research: What other solutions already exist? What are their limitations?
What basic direction is promising for a solution?
Empirical evaluation: how should the approach be empirically tested? What are the methods of comparison? What data sets, benchmarks, etc. should be used? What are the comparison metrics?

Registering the Thesis

As soon as the first examiner has confirmed and signed the assignment, the student can officially register the thesis at the examination office. To do this, a registration form from the examination office is required, in which it must first be confirmed that the student can in principle be admitted to the thesis. The second page should be filled out as far as possible and scanned to be sent afterward as a PDF by mail to the first examiner. The first examiner will take care of a second examiner and return the signed form. The secretariat will send everything by mail to the examination office with supervisor, first examiner, and student in CC. This will start the official process time.

Supervision

As a rule, the student and supervisor meet every two weeks to discuss the progress of the thesis. The supervisor has an advisory role; this means that the student should prepare the meeting independently and consider on which topics advice or assistance is currently needed. For example, it is a good idea to prepare a few slides to show the current progress of the thesis and to ask specific questions. The supervisor will make literature recommendations, share resources, and provide content advice, but will not directly work with code or write text. If needed, the supervisor may decide to include the primary investigator in these meetings. Outside of the meetings, the supervisor is generally available to the student if needed. However, we assume that most topics can be discussed in face-to-face meetings.

Concept Presentation

After 1-2 months of the official process time, a concept talk should be given at the institute to get feedback on the content orientation and schedule. The talk should last about 10-15 minutes with another 10 minutes for questions and discussion. As a rule, it will be in English. The presentation slides should be written using our slide template.

Evaluation

Bachelor Theses
We evaluate bachelor theses based on the structure of the thesis, linguistic quality, clarity of motivation and definition of the research question, description of the methodology, systematic experimental design, and interpretation of the results.

Structure: The thesis should first present the motivation and the necessary background knowledge. Then, the problem should be clearly and systematically defined to explain the actual research questions of the paper. The subsequent empirical evaluation should then answer these questions and identify future points of contact.

Language quality: The paper should be written in English. It should contain no errors, be written in a clear and understandable manner, and conform to scientific standards.

Motivation and background: The paper should motivate the problem by references to application and current research. The background presented should be concise but allow for an understanding of the rest of the paper.

Clarity of problem: The problem statement should be presented clearly and formally correct to clearly state the topic of the paper in a few research questions. The questions should be specific enough to be tested in the evaluation.

Methodology: Algorithms, concepts, and solution methods used to answer the research questions should be presented here in sufficient detail.

Experimental design: The empirical evaluation should be clearly described, scientifically accurate, and appropriate to the research questions. It should meet the standards of comparable research.

Interpretation of results: The evaluation criteria for the results should fit the questions of the paper. The results themselves should be adequately described and used to draw conclusions about the research questions.

At least half of these criteria must be met to successfully complete the paper. An exception is the linguistic quality: if frequent linguistic errors significantly affect the comprehensibility of the paper, this point alone can also lead to failure.
Master Theses
We evaluate bachelor theses based on the structure of the thesis, linguistic quality, clarity of motivation and definition of the problem, systematic literature analysis, description of methodology and novel approaches, systematic experimental design, interpretation of results, and independence during the process.

Structure: The paper should first present the motivation and the necessary background knowledge. Then, the problem should be clearly and systematically defined to explain the actual research questions of the thesis. The subsequent empirical evaluation should then answer these questions and identify future points of contact.

Language quality: The paper should be written in English. It should contain no errors, be written in a clear and understandable manner, and conform to scientific standards. The argumentation should be convincing throughout the paper. Illustrations should further emphasize comprehensibility.

Motivation and background: The paper should motivate the problem statement by references to application and current research. The background presented should be concise but allow for the understanding of the rest of the paper.

Literature review: Part of the paper should provide an overview of related research questions and approaches. The student should be able to independently identify and describe relevant work and explain its limitations.

Clarity of the problem statement: The problem statement should be clear and formally correct in order to clearly state the topic of the paper in a few research questions. The questions should be specific enough to be tested in the evaluation.

Methodology: Algorithms, concepts, and solution methods used to answer the research questions should be presented here in sufficient detail. Your own new ideas and extensions to existing approaches are expected.

Experimental design: The empirical evaluation should be clearly described, scientifically accurate, and appropriate to the research questions. It should meet the standards of comparable research.

Interpretation of results: The evaluation criteria for the results should fit the questions of the paper. The results themselves should be adequately described and used to draw conclusions about the research questions.

Independence: We expect that the majority of the thesis will be done independently; that is, the student should independently contribute new ideas and conduct experiments on their own, without too much assistance from the supervisor.

At least half of these criteria must be met in order to successfully complete the thesis. An exception is the linguistic quality: if frequent linguistic errors significantly impair the comprehensibility of the thesis, this point alone can also lead to failure.

Final presentation

Approximately two to four weeks after submission, the thesis is presented to the institute in an approximately 20-minute presentation. Afterward, another 20 minutes are planned for questions. As a rule, the final presentation should be given in English, just like the concept presentation. Our slide template should be used here.