Kategorie: ‘Theses’
Leveraging Large Language Models for Enhanced Decision Support in Home Energy Management
Large language models (LLMs) have proven the ability to assist diverse users in conducting a variety of individual tasks via intuitive and natural conversations. This thesis discusses a utilization of LLMs as a tool for informed decision-making in energy investments and operations. One major goal is to transform the way the consumers engage with home energy management systems and reduce expertise-related dependencies.
Generating security playbooks from attack-defense trees using Large Language Models
Applicability of Large Language Models for Evaluating Digital Exercises in Higher Education
As universities strive to enhance the effectiveness of their lecture exercises, there arises a need for diverse and realistic test user scenarios to evaluate the understandability and usefulness of educational materials. However, in creating such scenarios a number of challenges arise: Real world students can rarely be used for testing, they are likely inexperienced or experienced but have little incentive to refresh their knowledge to test tasks for lectures. In addition, leaking tasks to students can lead to skewed learning results during the actual course. On the other hand, simple unit tests can neglect the importance of a task being understandable to a human and do not offer much insight into task difficulty.
Knowledge Graph Construction from Biomedical Literature using Large Language Models
Fine-tuning pre-trained large language models (LLMs) enhances biomedical text mining. This thesis introduces a tool capable of performing tasks such as Named Entity Recognition (NER), Normalization (NEN), and Knowledge Graph Construction (KGC). A key research question explores how LLMs can address the challenges of named entity recognition, normalization, and relation extraction in biomedical contexts.
Supervised fine-tuning LLMs for cybersecurity playbook translation
Prompt Engineering for Translating unstructured playbooks into CACAO standard with the help of Large Language Models
A Mixed Reality Agent as a Navigation Guide
Mixed reality agents are simulated humans which can be viewed and interacted with on mixed reality technologies. For conveying content in educational settings, mixed reality agents have a series of advantages. They are part of a mixture of virtual elements with the real world, which provides opportunities where the agent can refer to points and locations in the real surroundings. Moreover, they are viewable in 3D from any angle so that an observer can freely choose the perspective to watch the agent. This is, e.g., not possible with videos where the viewpoint was chosen previously by the cameraman. The agents underlying behavior can also be made configurable and so, content can quickly be produced and changed as opposed to instructional videos where changes to the content require re-recording the video. In the educational sector, the combination of these factors means that mixed reality agents have a high potential to act as navigation guides. One issue that freshmen in universities face in the beginning is that they are unsure how to find the correct course room and are generally unaware of the building’s layout. Here, mixed reality agents can act as initial guides which inform students about their location and guide them around, e.g., pointing out where to find rooms. Additionally, they can add interesting information, e.g., about the organizational aspects, e.g., where important offices are or successful research projects of the different university chairs as a help in finding interests.
Frequent Learner Errors: NLP Insights into Automated Grading
This thesis aims to explore the realm of Natural Language Processing (NLP) in the context of automated grading systems, focusing specifically on identifying and analyzing frequent learner mistakes. With the increasing integration of technology in education, automated grading systems have gained prominence, but they often lack nuanced understanding and feedback provision. This research endeavors to enhance the efficacy of such a system in use at the institute by employing NLP techniques to dissect learner errors and provide more tailored feedback.
Automated Data Processing and Knowledge Discovery for Time Series Data utilizing Large Language Models
Large language models (LLMs) have proven the ability to assist diverse users in conducting a variety of individual tasks via intuitive and natural conversations. This thesis discusses a utilization of LLMs to perform (semi-)automated data processing and analyses on time series data. One major goal is to reduce expertise-related dependencies, allowing more people to manipulate data and gain beneficial insights.
Explainable Data – Trust, Transparency and Bias Mitigation in ML
This bachelor thesis aims to delve into the critical intersection of trust, transparency, and bias mitigation in machine learning (ML) systems through the lens of explainable data. The proliferation of ML algorithms across various domains has underscored the importance of understanding how these systems make decisions, especially when they impact individuals or societal outcomes.