Knowledge Graph Construction for Cultural Heritage Libraries

This technical report discusses the construction and refinement of knowledge graphs (KGs) aimed at enhancing digital libraries that house metadata for cultural heritage objects. It outlines the challenges faced in the representation, organization, and integration of diverse historical objects, emphasizing the need for improved metadata quality. The report presents a proposed ontology that ensures interoperability among various cultural heritage objects while maintaining specific semantics. Additionally, it details the application of information extraction techniques to address issues of incompleteness and inaccuracies in existing metadata. The document also highlights the significance of leveraging natural language processing (NLP) to enhance the quality of historical metadata and proposes methodologies for entity linking and knowledge graph refinement. The intended audience includes both domain experts and the general public, aiming to facilitate efficient search and retrieval of cultural heritage information.