본문 바로가기
  • Home

Semantic Data Processing of Sangseogohun and Sangseogoju, Part 2: Ontology Design and Parsing

  • Journal of Humanities, Seoul National University
  • 2025, 82(2), pp.219~268
  • DOI : 10.17326/jhsnu.82.2.202505.219
  • Publisher : Institute of Humanities, Seoul National University
  • Research Area : Humanities > Other Humanities
  • Received : April 10, 2025
  • Accepted : May 8, 2025
  • Published : May 31, 2025

Byeon Eunmi 1 Donghak Lee 1

1고려대학교

Accredited

ABSTRACT

This paper presents the second study on the semantic data processing of Sangseogohun (尙書古訓) and Sangseogoju (尙書古注). The study explores the process of converting semi-structured XML data into ontology-based triple data. Based on the structural characteristics of Confucian classics, as well as key contextual and interpretative information, this study designs an ontology focused on semantic relational links. A methodological framework is proposed to structurally analyze the interpretive logic and inter-commentary interactions embedded in Confucian texts, by means of data parsing and knowledge graph implementation. The conceptual structure encompasses textual, contextual and interpretative information, grounded in the shared design logic of XML schema and Ontology design. A total of 15 Classes are defined to capture the variations in data linkage between XML and Ontology. In particular, relationships among commentaries are categorized into three types — basic, nested, and dependent — to systematically reflect inter-citation patterns and the internal logical structures of commentaries. In addition, contextual information is linked through relation attributes to articulate semantic connectivity and interpretive lineages within the dataset. XML data are parsed using Python and LLMs to extract Nodes and Edges, thereby laying the foundation for building a graph database. Based on this, a partial knowledge graph is implemented to visualize the interpretive structure of Confucian texts, demonstrating the potential of semantic data modeling for knowledge integration and cross-textual comparison in Confucian studies.

Citation status

* References for papers published after 2023 are currently being built.