Application of Typed Operational System in Knowledge Graph Reasoning
Haipeng Peng
Zenodo (CERN European Organization for Nuclear Research)
Problems Identified (2)
Traditional KG reasoning limitations: Traditional knowledge graph reasoning methods have dispersed representations, difficulty with multi-path fusion, weak interpretability, and challenges integrating with deep learning.
Traditional KG reasoning limitations: Traditional knowledge graph reasoning methods have dispersed representations, difficulty with multi-path fusion, weak interpretability, and challenges integrating with deep learning.
Proposed Solutions (4)
Typed operational KG reasoning framework: The paper proposes a knowledge graph representation and reasoning framework based on a typed operational system that unifies entities, relations, and attributes as weighted sets.
Algebraic operation reasoning: The framework implements reasoning tasks using algebraic operations including union addition, Cartesian product, and projection.
Typed operational KG reasoning framework: The paper proposes a knowledge graph representation and reasoning framework based on a typed operational system that unifies entities, relations, and attributes as weighted sets.
Algebraic operation reasoning: The framework implements reasoning tasks using algebraic operations including union addition, Cartesian product, and projection.
Results (2)
Reasoning tasks implemented:
Reasoning tasks implemented:
Research Domain
Knowledge graph reasoning