Purpose: This study investigates the complex mappings from RDA to IFLA LRM, Dublin Core (DC), and MARC 21. It aims to identify key sources of structural and semantic complexity and inconsistency within these mappings, and to introduce indicators that quantify their consistency. The goal is to support a clearer understanding of mapping structures across different metadata schemas. Methods: Using RDA Vocabularies v5.1.0, we analyzed the mappings to the three target schemas through basic statistical summaries. We focused on two primary sources of complexity: the polyhierarchical nature of the RDA elements and the alignment of hierarchical or containment relationships between the source and target elements. To evaluate these aspects, we developed two sets of indicators: (a) element-based metrics and (b) tree-based metrics derived from the hierarchical structure of the RDA elements. Results: The mappings differed significantly in size and structural characteristics. (a) In all three maps, over 90% of the RDA element-to-target pairs maintained their structural consistency. When the consistency criteria were relaxed, the number of mismatches decreased, although some inconsistencies persisted. (b) Comparisons between RDA graph partitioning based on hierarchical trees and that based on bibliographic coupling in the target mappings revealed a moderate degree of similarity, as measured by element pair-based precision, recall, F-measure, and the Omega Index. These measures improved when target-side hierarchical relationships were incorporated and the root levels of the RDA trees were adjusted. Collectively, these indicators offered a comparative perspective on structural alignment and consistency across the three mapping sets.
Purpose: In the context of digital history, little has been done on the digitalization of the historical research process itself through acts of problematizing, collecting, interpreting, and narrating. However, enhancing the reproducibility and verification of these historical studies is crucial. Therefore, this study aims to design an ontology to structure the historical research process as data and demonstrate its usefulness through a case study involving actual historical research. Methods: First, based on the study done by two distinguished historians, we define historical research as a practice consisting of six activities: 'problematizing,' 'collecting,' 'source editing,' 'source interpreting,' 'colligating,' and 'narrating.' Then, by referring to the PROV Ontology, which is designed to describe the creation and history of things, we define the classes and their hierarchical relationships that express these six activities and the metadata requirements for each to establish specific vocabularies. This approach presents a new ontology that represents the historical research process based on individual activities in the research. Results: After constructing the data based on the proposed ontology using the author's historical research as a case study, SPARQL queries allow for the extraction of information, such as the relationship between specific historical narratives and the historical knowledge underpinning them, as well as the connections between that knowledge and the historical sources in which it is contained. This demonstrates that by constructing larger-scale data, it is possible to establish a digital infrastructure that enables the reproduction and verification of the research process.
