Inductive Knowledge Graph Completion with GNNs and Rules: An Analysis

arXiv:2308.07942v2 Announce Type: replace-cross
Abstract: The task of inductive knowledge graph completion requires models to learn inference patterns from a training graph, which can then be used to make predictions on a disjoint test graph. Rule-based methods seem like a natural fit for this task, but in practice they significantly underperform state-of-the-art methods based on Graph Neural Networks (GNNs), such as NBFNet. We hypothesise that the underperformance of rule-based methods is due to two factors: (i) implausible entities are not …

abstract analysis arxiv cs.ai cs.lg cs.si gnns graph inductive inference knowledge knowledge graph learn natural patterns practice predictions rules test training type