Abstract:Reasoning on knowledge graphs (KGs) is significant for downstream applications, such as question answering and information extraction. On the basis of using factual triples in KGs, learning logic rules has been explored to handle KG reasoning in an interpretable pattern. Some rule learning methods simply apply thresholds to reduce the exponential complexity caused by the vast candidate rules derived from KGs. Others are designed to prune candidates on the basis of embedding facts in the KG. However, these methods only consider the factual triples, which mainly contain structural information, and ignore sufficient semantic information in real-world KGs. This situation weakens the methods in searching for high-quality rules. To this end, we propose a joint method by incorporating logic rules with textual representations for interpretable KG reasoning, named LoTus. Firstly, LoTus obtains the structural embeddings with only factual triples, and the textual embeddings with descriptions of entities, which are later integrated into the representations of entities and relations. Secondly, top candidate rules are selected by a proposed joint pruning strategy incorporating structural and textual representations in KGs. Finally, LoTus captures high-quality rules and removes low-quality ones by a filtering process with metrics of rules. It reduces the time complexity on KG reasoning by learning rules and provides interpretability simultaneously. Extensive experiments on different datasets show that LoTus obtains competitive performance comparing to baselines on KG reasoning metrics Hits@1, Hits@10 and mean reciprocal rank (MRR), and achieves significant superiority among the recent rule learning methods.

Differentiable learning of rules with constants in knowledge graph

Differentiable Learning of Graph-like Logical Rules from Knowledge Graphs

Ruleformer: Context-aware Rule Mining over Knowledge Graph.

Iteratively Learning Embeddings and Rules for Knowledge Graph Reasoning

Rule mining over knowledge graphs via reinforcement learning

Learning Typed Rules over Knowledge Graphs

Ruleformer: Context-aware Differentiable Rule Mining over Knowledge Graph

A Hybrid Model for Learning Embeddings and Logical Rules Simultaneously from Knowledge Graphs

Knowledge Reasoning via Jointly Modeling Knowledge Graphs and Soft Rules

Towards Learning Instantiated Logical Rules from Knowledge Graphs

RulE: Neural-Symbolic Knowledge Graph Reasoning with Rule Embedding

A Comprehensive Study on Knowledge Graph Embedding over Relational Patterns Based on Rule Learning

Rule-enhanced Iterative Complementation for Knowledge Graph Reasoning

RulE: Knowledge Graph Reasoning with Rule Embedding

Differentiable Learning of Logical Rules for Knowledge Base Reasoning

ChatRule: Mining Logical Rules with Large Language Models for Knowledge Graph Reasoning

Missing-edge Aware Knowledge Graph Inductive Inference Through Dual Graph Learning and Traversing

TILP: Differentiable Learning of Temporal Logical Rules on Knowledge Graphs

Incorporating Logic Rules with Textual Representations for Interpretable Knowledge Graph Reasoning

Enabling Abductive Learning to Exploit Knowledge Graph

RLogic: Recursive Logical Rule Learning from Knowledge Graphs