SP
BravenNow
Knowledge Graphs Generation from Cultural Heritage Texts: Combining LLMs and Ontological Engineering for Scholarly Debates
| USA | technology | ✓ Verified - arxiv.org

Knowledge Graphs Generation from Cultural Heritage Texts: Combining LLMs and Ontological Engineering for Scholarly Debates

#knowledge graphs #cultural heritage #Large Language Models #ontological engineering #scholarly debates #semantic analysis

📌 Key Takeaways

  • The article explores generating knowledge graphs from cultural heritage texts using a combination of Large Language Models (LLMs) and ontological engineering.
  • It focuses on applying this methodology to support and enhance scholarly debates by structuring complex information.
  • The approach aims to improve data interoperability and semantic understanding in cultural heritage research.

📖 Full Retelling

arXiv:2511.10354v1 Announce Type: cross Abstract: Cultural Heritage texts contain rich knowledge that is difficult to query systematically due to the challenges of converting unstructured discourse into structured Knowledge Graphs (KGs). This paper introduces ATR4CH (Adaptive Text-to-RDF for Cultural Heritage), a systematic five-step methodology for Large Language Model-based Knowledge Extraction from Cultural Heritage documents. We validate the methodology through a case study on authenticity

🏷️ Themes

Digital Humanities, Semantic Technologies

Entity Intersection Graph

No entity connections available yet for this article.

Deep Analysis

Why It Matters

This research matters because it bridges the gap between traditional humanities scholarship and modern computational methods, potentially revolutionizing how historians and cultural researchers access and analyze primary sources. It affects academic institutions, cultural heritage organizations, and digital humanities researchers by providing new tools for knowledge discovery. The methodology could democratize access to historical analysis and enable more sophisticated interrogation of cultural texts than previously possible.

Context & Background

  • Cultural heritage institutions worldwide have been digitizing collections for decades, creating massive archives of unstructured text
  • Traditional knowledge graph construction often requires extensive manual annotation by domain experts
  • Large Language Models have shown remarkable capability in understanding and processing natural language
  • Ontological engineering provides formal structures for representing domain knowledge in machine-readable formats
  • Digital humanities has emerged as an interdisciplinary field combining computational methods with traditional humanities research

What Happens Next

Researchers will likely refine their methodology through case studies with specific cultural heritage collections, publish validation results in academic journals, and potentially develop open-source tools for wider adoption. Cultural institutions may begin pilot projects to apply this approach to their digitized collections, and funding agencies might support further development of these hybrid AI-humanities methodologies.

Frequently Asked Questions

What are knowledge graphs in this context?

Knowledge graphs are structured representations of information that connect entities, concepts, and relationships extracted from cultural heritage texts, creating a semantic network that machines can query and analyze.

How do LLMs and ontological engineering work together?

LLMs process and extract information from unstructured text, while ontological engineering provides the formal framework to structure this information into meaningful relationships and categories that reflect scholarly understanding of the domain.

What types of cultural heritage texts could this approach handle?

This methodology could potentially process historical documents, manuscripts, letters, diaries, archival records, and other textual artifacts from cultural heritage collections that have been digitized but remain largely unstructured.

How does this benefit scholarly debates?

By structuring information into queryable knowledge graphs, researchers can more easily identify patterns, connections, and contradictions across large corpora, enabling more evidence-based scholarly discussions and new research questions.

What are the main challenges in this approach?

Key challenges include ensuring historical accuracy, managing ambiguity in historical language, addressing biases in training data, and creating ontologies that adequately represent complex historical and cultural concepts.

}
Original Source
arXiv:2511.10354v1 Announce Type: cross Abstract: Cultural Heritage texts contain rich knowledge that is difficult to query systematically due to the challenges of converting unstructured discourse into structured Knowledge Graphs (KGs). This paper introduces ATR4CH (Adaptive Text-to-RDF for Cultural Heritage), a systematic five-step methodology for Large Language Model-based Knowledge Extraction from Cultural Heritage documents. We validate the methodology through a case study on authenticity
Read full article at source

Source

arxiv.org

More from USA

News from Other Countries

🇬🇧 United Kingdom

🇺🇦 Ukraine