Abschlussarbeit

Beschreibung:

Retrieval-Augmented Generation (RAG) systems have shown significant potential for domain-specific Question Answering (QA) tasks, although persistent challenges in retrieval precision and context selection continue to hinder their effectiveness. This thesis investigates the integration of Knowledge Graphs (KGs) and Retrieval-Augmented Generation (RAG) to enhance the performance of large language model (LLM)-based chatbots in processing technical product documentation. It begins with a theoretical foundation of both RAG and KGs, followed by a comparative analysis of existing KG construction approaches, highlighting their methodologies, strengths, and limitations. The core objective is to examine how structured knowledge—whether used independently or in combination with retrieval-based methods—affects the factual accuracy, completeness, and reasoning quality of chatbot-generated responses. To this end, a modular experimental setup will be developed, allowing for the evaluation of different system variants: LLM with RAG, LLM with KGs, and a hybrid LLM-RAG-KG approach.