Application of Typed Operational System in Knowledge Graph Reasoning
Haipeng Peng
Zenodo (CERN European Organization for Nuclear Research)
Problems Identified (2)
Limitations of traditional KG reasoning: Traditional knowledge graph reasoning methods have dispersed representation, difficulty fusing multi-path information, weak interpretability, and poor integration with deep learning.
Limitations of traditional KG reasoning: Traditional knowledge graph reasoning methods have dispersed representation, difficulty fusing multi-path information, weak interpretability, and poor integration with deep learning.
Proposed Solutions (4)
Typed operational KG reasoning framework: The paper proposes a typed operational system framework that represents and reasons over knowledge graphs by unifying entities, relations, and attributes as weighted sets.
Algebraic KG reasoning operations: The approach implements reasoning tasks using algebraic operations such as union addition, Cartesian product, and projection.
Typed operational KG reasoning framework: The paper proposes a typed operational system framework that represents and reasons over knowledge graphs by unifying entities, relations, and attributes as weighted sets.
Algebraic KG reasoning operations: The approach implements reasoning tasks using algebraic operations such as union addition, Cartesian product, and projection.
Results (2)
KG reasoning task implementation:
KG reasoning task implementation:
Research Domain
Knowledge graph reasoning