Ask Dataverse and Large Language Models in the RAG framework
Year: 2025
Type: article
Abstract: Knowledge graphs and Large Language Models (LLMs) offer complementary strengths in AI-driven reasoning and decision-making. While knowledge graphs are dynamic, human-curated, and distributed within the Semantic Web, LLMs are centralized, static, and machine-driven. This contrast presents both challenges and opportunities for integration. By integrating LLMs with knowledge graphs, we can enhance contextual understanding, improve reasoning, and introduce temporal awareness—ultimately creating more explainable and adaptable AI systems. The integration framework called Ghostwriter combines both LLMs and knowledge graphs and provides access to any type of content in the custom curated collections, such as web pages, documentation, guides, Google Threads, etc.
Author: Tykhonov, Vyacheslav
Institution: Royal Netherlands Academy of Arts and Sciences
Language: English
Topic: Topic Modeling
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W6949445088
FWCI:
Cites:
Cited by:
Related to: