A Conceptual Framework for Inclusive Semantic Retrieval in Digital Heritage Archives KianLam Tan HengWei Lee ChenKim Lim
School of Digital Technology, Wawasan Open University, George Town 10050, Penang, Malaysia
Institute for Environment and Development (LESTARI), Universiti Kebangsaan Malaysia (UKM), Bangi, Malaysia
Abstract
Cultural heritage platforms such as Europeana face the challenge of providing equal access to linked data collections to the widest possible audience. Semantic search tools of large-scale heritage data are at the moment not very intuitive for the average non-expert user. This conceptual proposal paper makes the case for the development and design of a generic model to better facilitate the discovery and retrieval of information in cultural heritage knowledge graphs. It presents the results of a preliminary mapping of information needs and a framework to enhance information retrieval performance for linked data models, in particular, to the Europeana Data Model (EDM) by adding semantic weighted relationships between the various entities described by the RDF schema. It combines approaches from category-based weighting of graph edges (incorporating contextual semantic information related to entity relationships in the knowledge graph), the exploitation of external linked data sources such as DBpedia and Freebase and graph traversal algorithms for path ranking, to enable more accurate and inclusive search results based on user queries. The proposed model can support non-expert users, for example educators and historically underrepresented communities, to explore cultural artifacts, potentially even if no exact match for search terms can be found in the data or no specific domain knowledge exists. This aligns with an agenda for inclusive digital communication which goes beyond access to data, to its proactive semantically-mediated discovery. The paper concludes with a plan for future evaluation and the potential for cross-domain application to digital libraries, and cross-cultural linked data collections.
Keywords: Digital Cultural Heritage, Semantic Search, Weighted Relationship, Semantic Search, Linked Open Data
Topic: Digital Communication for Inclusive Societies
