Software Engineering

Large Language Models (LLMs) are increasingly used to create educational content such as quizzes. While generation quality has improved, there is no standardized, reproducible benchmark for evaluating assessments against pedagogically relevant criteria such as difficulty, fidelity to the source material, coverage, and distractor quality.

This thesis constructs and validates a modular benchmarking framework to systematically and reproducibly evaluate quizzes using multiple LLMs as judges. By providing structured, rubric-driven scores and logging all evaluation details, the framework quantifies variance, supports robust aggregation, and produces actionable benchmarking reports. This system provides a systematic, reproducible approach to evaluating quizzes against pedagogically relevant criteria.

Developing high-quality quiz questions within Artemis currently necessitates significant manual effort, deep domain expertise, and strict alignment with course contents. Consequently, editors face challenges in maintaining robust question pools, often resulting in limited practice material for students. This thesis proposes integrating generative artificial intelligence (AI) into the Artemis platform to streamline the quiz creation lifecycle.

The proposed solution establishes a human-in-the-loop workflow where editors define specific constraints, such as topic and question count, to produce structured initial drafts. Furthermore, generation quality is enhanced by utilizing internal platform data like course content and learning competencies. Finally, a dedicated refinement layer empowers editors to iteratively adjust drafts via natural language instructions prior to final approval.

Doctoral application portals manage admissions and shape candidate experience. Many systems create barriers through complex interfaces and inconsistent design. TUMApply at the Technical University of Munich replaces an earlier system but lacks a clear, inclusive, and consistent interface that supports efficient task completion and equitable participation.

This thesis applies Nielsen’s usability heuristics with gender-sensitive and accessible design practices to identify and address usability barriers and inclusivity weaknesses. The primary contribution is a redesigned interface that demonstrates how doctoral application portals can better serve diverse user needs while maintaining efficiency. This work offers practical insights for building academic platforms that support usability and inclusivity, grounded in systematic analysis, iterative design, and implementation.

This thesis develops an analog story-coding activity using Ozobot for German elementary school education. The activity combines storytelling with hands-on robotics to introduce fundamental programming concepts and computational thinking skills. Pupils guide the robot through stories using color-coded cards, without the need for text- or codeblock-based programming. Through iterative testing, the thesis examines the impact of this approach on pupils’ motivation, the activity’s usability and it’s feasibility in real teaching conditions. The thesis summarizes the findings to provide recommendations for designing story-driven analog robotics activities for elementary schools.

This Master’s thesis presents the design and implementation of a digital story-coding activity for elementary education using Ozobots. It combines narrative elements with block-based programming to introduce core computational thinking concepts in an engaging and accessible way. The study develops a customized Ozoblockly environment in German, reducing technical barriers and preparation effort for teachers. The story-driven approach makes programming tangible and expressive, fostering creativity, motivation, and inclusive participation. User testing in classrooms examines pupils’ engagement, intuitive use of programming concepts, and classroom feasibility, providing formative insights to guide iterative improvements. The results inform the design of low-threshold, inclusive computational thinking materials for early education.

This bachelor’s thesis aims to advance the Ozobot Board Game by transforming an existing prototype of an educational board game into a sustainable and classroom-ready product. The project focuses on developing a durable, engaging, and easy-to-use tool for teaching basic programming concepts. It seeks to enhance the game’s replayability while promoting equal participation and interest in computing education among all children, regardless of gender or cultural background. The design includes differentiated learning tasks, dynamic difficulty adjustment, and iterative refinement through user feedback and classroom testing. The final outcome will be a classroom-ready board game package, including adaptable learning resources and a teacher’s manual, designed for inclusive and flexible use in diverse educational settings.

Recruiting doctoral candidates requires clear visibility of relevant research opportunities and efficient support throughout the application process. As TUMApply evolves into a unified portal for managing doctoral applications at the Technical University of Munich, further enhancements are necessary to strengthen usability and engagement for prospective candidates. This thesis designs and implements a topic-based job categorization and notification system that enables applicants to subscribe to subject areas and receive automated email alerts when relevant positions become available. A dedicated profile section allows users to manage personal data and upload reusable documents, improving usability and reducing repetitive effort. The project further ensures that each submitted application remains exactly as it was at the moment of submission, so later updates to a user’s profile no longer alter previously completed applications.

Competency-based educational recommender systems enable personalized learning, but the absence of standardized benchmark datasets prevents systematic algorithmic comparison. This thesis develops a platform for collaborative competency relationship data collection with four objectives: a generalized database model, authentication systems for credential verification, administrative interfaces for dataset management and export, and production deployment infrastructure. The system enables systematic collection of high-quality competency mapping data for creating reliable benchmarks.

Complex writing tasks increasingly rely on artificial intelligence to support in domains requiring regulatory compliance and clarity. In academic recruitment, the wording of job advertisements is critical for ensuring transparency, fairness, and equal opportunity. TUMApply currently requires professors to manually write job advertisements, often resulting in inconsistent quality, missing details, and gender-coded or non-compliant language. This thesis integrates an AI assistant into the workflow to automate draft generation, flag biased or legally problematic phrasing, and suggest improvements, reducing effort while enhancing clarity, inclusiveness, and compliance.

This thesis focuses on transitioning an experimental prototype to a production-ready Open Educational Resources (OER) platform. LEARN-Hub aims to distribute standardized teaching materials for computer science (CS) education. The existing system requires increased maintenance effort and offers limited support for processing teaching materials at scale.

The work migrates the server from Flask to Spring Boot to align the platform with institutional standards. It designs a pipeline to transform unstructured PDFs into structured data for consistent visualization. In addition, the work refines the React client to adhere to established usability heuristics and to ensure coherent presentation of the generated content.

TUMApply, the doctoral application portal of the Technical University of Munich, centralizes applications and job postings for a transparent process. However, manual data handling currently leads to inconsistencies, high administrative workloads and reduced clarity for both applicants and professors. This thesis extends TUMApply by introducing AI-supported features that enhance usability. For applicants, the system implements automated data extraction from uploaded documents to prefill profiles and reduce redundant entry. For professors, the system provides AI-based translation suggestions for professors and an enhanced review module to ensure job postings meet all criteria before publication.