2023/09/07 by J.H. Cobb, Gebhart, Thomas, Thomas Gebhart +1
Computer Science · #Advanced Graph Neural Networks #Artificial Intelligence (cs.AI) #Artificial intelligence #Computer science #Embedding #Explainable Artificial Intelligence (XAI) #FOS: Computer and information sciences #Graph #Graph embedding #Inference #Information Retrieval (cs.IR) #Interpretability #Machine learning #Scalability #Social and Information Networks (cs.SI) #Theoretical computer science #Topic Modeling #Transduction (biophysics)
paper · pdf · doi:10.48550/arxiv.2309.03773
openalex publication_date 2023/09/07 · openalex created_date 2023/09/09 · openalex updated_date 2026/07/28
Many inference tasks on knowledge graphs, including relation prediction, operate on knowledge graph embeddings -- vector representations of the vertices (entities) and edges (relations) that preserve task-relevant structure encoded within the underlying combinatorial object. Such knowledge graph embeddings can be modeled as an approximate global section of a cellular sheaf, an algebraic structure over the graph. Using the diffusion dynamics encoded by the corresponding sheaf Laplacian, we optimally propagate known embeddings of a subgraph to inductively represent new entities introduced into the knowledge graph at inference time. We implement this algorithm via an efficient iterative scheme and show that on a number of large-scale knowledge graph embedding benchmarks, our method is competitive with -- and in some scenarios outperforms -- more complex models derived explicitly for inductive knowledge graph reasoning tasks.