{"id":98,"date":"2021-11-16T07:42:20","date_gmt":"2021-11-16T07:42:20","guid":{"rendered":"http:\/\/muenter.informatik.rwth-aachen.de\/dbis\/?page_id=98"},"modified":"2024-11-19T15:16:58","modified_gmt":"2024-11-19T14:16:58","slug":"thesis-projects","status":"publish","type":"page","link":"https:\/\/dbis.rwth-aachen.de\/dbis\/index.php\/academics\/thesis-projects\/","title":{"rendered":"Thesis Projects"},"content":{"rendered":"\n<h4 class=\"wp-block-heading has-text-align-center has-text-color has-background\" style=\"background-color:#d7ffed\"><a href=\"https:\/\/dbis.rwth-aachen.de\/dbis\/index.php\/academics\/thesis-projects\/information-about-thesis-process\/\">Information<\/a>&nbsp;about Diploma\/Master thesis process<\/h4>\n\n\n\n<h4 class=\"wp-block-heading\" id=\"open-theses\">Open Theses<\/h4>\n\n\n\n<p class=\"wp-block-paragraph\"> <ul class=\"lcp_catlist\" id=\"lcp_instance_0\"><li><h4 class=\"lcp_title\"><a href=\"https:\/\/dbis.rwth-aachen.de\/dbis\/index.php\/2026\/a-generative-foundation-model-for-knowledge-graphs-geometry-and-text-aware-pretraining-for-transferable-downstream-tasks\/\">A Generative Foundation Model for Knowledge Graphs: Geometry- and Text-Aware Pretraining for Transferable Downstream Tasks<\/a><\/h4><hr class=\"lcp_excerpt\">\nKnowledge graphs like Wikidata combine rich relational structure with natural-language descriptions, yet most models are trained narrowly for a single task and transfer poorly. This thesis investigates how a single generative graph foundation model, pretrained on large-scale text-rich knowledge graphs, can be adapted to a range of downstream tasks, including knowledge graph completion, text-conditional subgraph ...<\/hr><\/li><li><h4 class=\"lcp_title\"><a href=\"https:\/\/dbis.rwth-aachen.de\/dbis\/index.php\/2026\/comparative-analyses-of-hybrid-llms-with-knowledge-base-integration-and-rags-in-biomedical-domain\/\">Comparative analyses of hybrid LLMs with Knowledge base integration and RAGs in biomedical domain<\/a><\/h4><hr class=\"lcp_excerpt\">\nLarge language models (LLMs) are increasingly used in biomedical applications, including literature mining (PMID: 40188094), drug discovery (PMID: 38730226; 41362614; https:\/\/arxiv.org\/abs\/2510.27130), clinical decision support (PMID: 40753316), and patient data analysis (PMID: 41034564). Hybrid approaches combining LLMs with structured knowledge bases and retrieval-augmented generation (RAG) improve performance and interpretability (PMID: 38830083; https:\/\/www.biorxiv.org\/content\/10.1101\/2025.05.08.652829v2) . However, LLM-based systems ...<\/hr><\/li><\/ul>  <\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><\/p>\n\n\n\n<h4 class=\"wp-block-heading\" id=\"running-theses\">Running Theses<\/h4>\n\n\n\n<p class=\"wp-block-paragraph\"> <ul class=\"lcp_catlist\" id=\"lcp_instance_0\"><li><h4 class=\"lcp_title\"><a href=\"https:\/\/dbis.rwth-aachen.de\/dbis\/index.php\/2026\/cross-schema-transferable-graph-learning-on-relational-databases-master-thesis\/\">Cross-Schema Transferable Graph Learning on Relational Databases (Master Thesis)<\/a><\/h4><hr class=\"lcp_excerpt\">\nRelational Deep Learning (RDL) enables predictive learning directly on relational databases by representing tables and foreign-key relationships as heterogeneous graphs. RelBench provides a benchmark for evaluating such models across databases with different schemas and downstream prediction tasks , can model typed relationships, but their schema-dependent representations ...<\/hr><\/li><li><h4 class=\"lcp_title\"><a href=\"https:\/\/dbis.rwth-aachen.de\/dbis\/index.php\/2026\/budget-aware-language-graph-pretraining-master-thesis\/\">Budget-Aware Language\u2013Graph Pretraining (Master Thesis)<\/a><\/h4><hr class=\"lcp_excerpt\">\nKnowledge-augmented question answering (QA) combines language understanding with relational evidence from knowledge graphs (KGs). QA-GNN  learns joint representations through masked language modeling and KG link prediction.\n\n\nHowever, retrieved evidence can exceed practical text and graph budgets. Fixed-size inputs and sampling or truncation may discard useful relational ...<\/hr><\/li><li><h4 class=\"lcp_title\"><a href=\"https:\/\/dbis.rwth-aachen.de\/dbis\/index.php\/2026\/kg-grounded-language-graph-models-for-task-agnostic-question-answering-master-thesis\/\">KG-Grounded Language\u2013Graph Models for Task-Agnostic Question Answering (Master Thesis)<\/a><\/h4><hr class=\"lcp_excerpt\">\nQA-GNN  motivate candidate-defined answer spaces.\n\n\nYet joint representation pretraining does not directly establish a reusable candidate-decision function. In particular, it remains unclear ...<\/hr><\/li><li><h4 class=\"lcp_title\"><a href=\"https:\/\/dbis.rwth-aachen.de\/dbis\/index.php\/2026\/towards-a-temporal-heterogeneous-graph-foundation-model-for-academic-knowledge-graphs-master-thesis\/\">Towards a Temporal Heterogeneous Graph Foundation Model for Academic Knowledge Graphs (Master Thesis)<\/a><\/h4><hr class=\"lcp_excerpt\">\nThis thesis investigates how to develop a foundation-model-style representation learner for Academic Knowledge Graphs (AKGs), such as arXiv-, DBLP-, MAG-, and Open Academic Graph-derived data. Academic graphs connect papers, authors, venues, institutions, fields of study, keywords, and citations through typed relations that evolve over time, making them a natural setting for temporal heterogeneous graph representation ...<\/hr><\/li><li><h4 class=\"lcp_title\"><a href=\"https:\/\/dbis.rwth-aachen.de\/dbis\/index.php\/2026\/composable-source-descriptions-and-declarative-reasoning-for-document-grounded-knowledge-graph-validation\/\">Composable Source Descriptions and Declarative Reasoning for Document-Grounded Knowledge Graph Validation<\/a><\/h4><hr class=\"lcp_excerpt\">\nKnowledge graphs built automatically from documents inherit extraction errors: values get truncated, entities are dropped, or records are bound to the wrong document. Structural validators such as SHACL catch formatting violations, but they cannot answer the question that matters for curation: does the source document actually contain the fact the graph claims?\n\n<\/hr><\/li><\/ul>   <\/p>\n\n\n\n<p class=\"has-text-align-right wp-block-paragraph\"><a href=\"https:\/\/dbis.rwth-aachen.de\/dbis\/index.php\/academics\/thesis-projects\/running-theses\/\">View all running theses<\/a><\/p>\n\n\n\n<h4 class=\"wp-block-heading\" id=\"completed-theses\">Completed Theses<\/h4>\n\n\n\n<p class=\"wp-block-paragraph\"> <ul class=\"lcp_catlist\" id=\"lcp_instance_0\"><li><h4 class=\"lcp_title\"><a href=\"https:\/\/dbis.rwth-aachen.de\/dbis\/index.php\/2026\/developing-an-explainable-anomaly-detection-system-forsmart-grids-by-incorporating-structural-and-operational-gridknowledge\/\">Developing an Explainable Anomaly Detection System forSmart Grids by Incorporating Structural and Operational GridKnowledge<\/a><\/h4><hr class=\"lcp_excerpt\">\nThesis Type\n\n\nMaster\n\n\nStudent: Sebastian Miller\n\n\nStatus\n\n\nIn Progress\n\n\nBackground\n\n\nSupervisory control and data acquisition (SCADA) systems are increasingly connected through information and communication technologies, exposing smart grids to cyberattacks and operational disruptions. Conventional signature-based intrusion detection systems (IDSs) reliably identify known attacks but cannot detect previously unseen patterns, while statistical and machine-learning-based IDSs may achieve high detection rates but often ...<\/hr><\/li><li><h4 class=\"lcp_title\"><a href=\"https:\/\/dbis.rwth-aachen.de\/dbis\/index.php\/2026\/a-comparison-of-interaction-modalities-for-extended-reality-agents\/\">A Comparison of Interaction Modalities for Extended Reality Agents<\/a><\/h4><hr class=\"lcp_excerpt\">\nCurrently, the primary way users interact with Large Language Models (LLMs) is through two-dimensional chat interfaces. However, for use cases in Extended Reality (XR) environments, the interaction paradigm shifts from a flat screen to a spatial experience. Here, LLMs can, e.g., be represented as XR agents, a personified version of the LLM. While 3D environments ...<\/hr><\/li><li><h4 class=\"lcp_title\"><a href=\"https:\/\/dbis.rwth-aachen.de\/dbis\/index.php\/2026\/traceability-framework-for-human-llm-assisted-tabular-data-transformations\/\">Traceability Framework for Human\u2013LLM-Assisted Tabular Data Transformations<\/a><\/h4><hr class=\"lcp_excerpt\">\nLarge Language Models (LLMs) are increasingly used to support data wrangling, but their integration into interactive transformation workflows raises new challenges for auditability, reproducibility, and accountability. When users approve, reject, or refine LLM-generated suggestions, conventional data lineage systems often fail to capture why a change occurred, who was responsible for it, and which transformation produced ...<\/hr><\/li><li><h4 class=\"lcp_title\"><a href=\"https:\/\/dbis.rwth-aachen.de\/dbis\/index.php\/2026\/training-a-tiny-llm-with-block-attention-residuals-on-commonsenseqa\/\">Training a Tiny LLM with Block Attention Residuals on CommonsenseQA<\/a><\/h4><hr class=\"lcp_excerpt\">\nKnowledge-augmented multiple-choice question answering (MCQA) aims to improve robustness and factual grounding by integrating external structured knowledge (e.g., knowledge graphs) into language-model-based decision making. Current high-performing systems typically retrieve a local subgraph relevant to a question and candidate answers, then combine pretrained language representations with explicit graph reasoning modules.\n\n\nThis thesis investigates an alternative representation path: ...<\/hr><\/li><li><h4 class=\"lcp_title\"><a href=\"https:\/\/dbis.rwth-aachen.de\/dbis\/index.php\/2026\/from-triples-to-pixels-visual-knowledge-graph-encoding-for-knowledge-augmented-multiple-choice-question-answering\/\">From Triples to Pixels: Visual Knowledge Graph Encoding for Knowledge-Augmented Multiple-Choice Question Answering<\/a><\/h4><hr class=\"lcp_excerpt\">\nKnowledge-augmented multiple-choice question answering (MCQA) aims to improve robustness and factual grounding by integrating external structured knowledge (e.g., knowledge graphs) into language-model-based decision making. Current high-performing systems typically retrieve a local subgraph relevant to a question and candidate answers, then combine pretrained language representations with explicit graph reasoning modules.\n\n\nThis thesis investigates an alternative representation path: ...<\/hr><\/li><\/ul>   <\/p>\n\n\n\n<p class=\"has-text-align-right wp-block-paragraph\"><a href=\"https:\/\/dbis.rwth-aachen.de\/dbis\/index.php\/academics\/thesis-projects\/completed-theses\/\" data-type=\"URL\" data-id=\"https:\/\/dbis.rwth-aachen.de\/dbis\/index.php\/academics\/thesis-projects\/completed-theses\/\">View all completed theses<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Information&nbsp;about Diploma\/Master thesis process Open Theses Running Theses View all running theses Completed Theses View all completed theses<\/p>\n","protected":false},"author":2,"featured_media":0,"parent":121,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"_acf_changed":false,"footnotes":""},"class_list":["post-98","page","type-page","status-publish","hentry"],"acf":[],"_links":{"self":[{"href":"https:\/\/dbis.rwth-aachen.de\/dbis\/index.php\/wp-json\/wp\/v2\/pages\/98","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/dbis.rwth-aachen.de\/dbis\/index.php\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/dbis.rwth-aachen.de\/dbis\/index.php\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/dbis.rwth-aachen.de\/dbis\/index.php\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/dbis.rwth-aachen.de\/dbis\/index.php\/wp-json\/wp\/v2\/comments?post=98"}],"version-history":[{"count":15,"href":"https:\/\/dbis.rwth-aachen.de\/dbis\/index.php\/wp-json\/wp\/v2\/pages\/98\/revisions"}],"predecessor-version":[{"id":5635,"href":"https:\/\/dbis.rwth-aachen.de\/dbis\/index.php\/wp-json\/wp\/v2\/pages\/98\/revisions\/5635"}],"up":[{"embeddable":true,"href":"https:\/\/dbis.rwth-aachen.de\/dbis\/index.php\/wp-json\/wp\/v2\/pages\/121"}],"wp:attachment":[{"href":"https:\/\/dbis.rwth-aachen.de\/dbis\/index.php\/wp-json\/wp\/v2\/media?parent=98"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}