Rosetta Statements: Lowering the Barrier for Semantic Parsing and Increasing the Cognitive Interoperability of Knowledge Graphs

• DOI: 10.48550/arxiv.2407.20007
• Publication Date: 2024-07-29
• Authors: Lars Vogt; Marcel Konrad; Kheir Eddine Farfar; Manuel Prinz; Allard Oelen
• Keywords: FOS: Computer and information sciences; Databases (cs.DB)

Machines need data and metadata to be machine-actionable FAIR (findable, accessible, interoperable, reusable) manage increasing volumes. Knowledge graphs ontologies are key this, but their use is hampered by high access barriers due required prior knowledge in semantics modelling. The Rosetta Statement approach proposes modeling English natural language statements instead of a mind-independent reality. We propose metamodel for creating semantic schema patterns simple statement types. supports versioning provides detailed editing history. Each pattern has dynamic label displaying as sentences. Implemented the Open Research Graph (ORKG) case, this allows domain experts define without needing knowledge. Future plans include combining Statements with units organize ORKG into meaningful subgraphs, improving usability. A search interface querying SPARQL or Cypher also planned, along tools entry display using Large Language Models NLP. two-step graph construction procedure. Domain can model content support from ontology engineers, lowering cognitive interoperability. second level involves developing reasoning, requiring collaboration engineers.