DualRep: Knowledge Graph Completion by Utilizing Dual Representation of Relational Paths and Tail Node Density Insights

Abstract

Knowledge graphs (KGs) possess a vital role in enhancing the semantic comprehension of extensive datasets across many fields. It facilitate activities like recommendation systems, semantic searching, and intelligent data mining. However, lacking information can sometimes limit the usefulness of knowledge graphs (KGs), as the lack of relationships between entities could severely limit their practical application. Most existing approaches for KG completion primarily concentrate on embedding-based methods or just use relational paths, neglecting the valuable structural information offered by node density. This research presents an approach that effectively combines relational paths and the density features of tail nodes to enhance the accuracy of predicting relationships that are missing in knowledge graphs. Our method combines the sequential relational context represented by paths with the structural prominence indicated by node density, allowing for a dual view on possible entity connections. We validate the effectiveness of our technique by conducting comprehensive tests on many benchmark datasets, revealing substantial enhancements compared to conventional approaches. The Dual-Rep model, which incorporates relational paths and node density features, has continuously shown improved performance across several metrics, such as Mean Reciprocal Rank (MRR), Hit at 1 (Hit@1), and Hit at 3 (Hit@3). The DualRep model achieved a mean reciprocal rank (MRR) of 90.80. Additionally, it achieved a hit rate of 87.39 at rank 1 (Hit@1) and a hit rate of 91.18.

Keywords

• Knowledge graph completion
• relational paths
• node density analysis
• graph structural features
• entity relationship prediction
• graph neural networks