Abstract:Knowledge graph reasoning (KGR) aims to infer new knowledge or detect noises, which is essential for improving the quality of knowledge graphs. Recently, various KGR techniques, such as symbolic- and embedding-based methods, have been proposed and shown strong reasoning ability. Symbolic-based reasoning methods infer missing triples according to predefined rules or ontologies. Although rules and axioms have proven effective, it is difficult to obtain them. Embedding-based reasoning methods represent entities and relations as vectors, and complete KGs via vector computation. However, they mainly rely on structural information and ignore implicit axiom information not predefined in KGs but can be reflected in data. That is, each correct triple is also a logically consistent triple and satisfies all axioms. In this paper, we propose a novel NeuRal Axiom Network (NeuRAN) framework that combines explicit structural and implicit axiom information without introducing additional ontologies. Specifically, the framework consists of a KG embedding module that preserves the semantics of triples and five axiom modules that encode five kinds of implicit axioms. These axioms correspond to five typical object property expression axioms defined in OWL2, including ObjectPropertyDomain, ObjectPropertyRange, DisjointObjectProperties, IrreflexiveObjectProperty and AsymmetricObjectProperty. The KG embedding module and axiom modules compute the scores that the triple conforms to the semantics and the corresponding axioms, respectively. Compared with KG embedding models and CKRL, our method achieves comparable performance on noise detection and triple classification and achieves significant performance on link prediction. Compared with TransE and TransH, our method improves the link prediction performance on the Hits@1 metric by 22.0% and 20.8% on WN18RR-10% dataset, respectively.

Probing Structural Knowledge from Pre-trained Language Model for Argumentation Relation Classification

Hi-ArG: Exploring the Integration of Hierarchical Argumentation Graphs in Language Pretraining

DMON: A Simple yet Effective Approach for Argument Structure Learning

ReasoningLM: Enabling Structural Subgraph Reasoning in Pre-trained Language Models for Question Answering over Knowledge Graph

Neural Axiom Network for Knowledge Graph Reasoning

A Joint Framework for Argumentative Text Analysis Incorporating Domain Knowledge

Rethinking Relation Classification with Graph Meaning Representations

Seeing both sides: context-aware heterogeneous graph matching networks for extracting-related arguments

GUNDAM: Aligning Large Language Models with Graph Understanding

Unifying Structure Reasoning and Language Pre-Training for Complex Reasoning Tasks

Hierarchical Neural Network: Integrate Divide-and-Conquer and Unified Approach for Argument Unit Recognition and Classification

Graph Structure Enhanced Pre-Training Language Model for Knowledge Graph Completion

Exploring fine-grained syntactic information for aspect-based sentiment classification with dual graph neural networks

Unifying Structure Reasoning and Language Model Pre-training for Complex Reasoning

Here's My Point: Joint Pointer Architecture for Argument Mining

Debate on Graph: a Flexible and Reliable Reasoning Framework for Large Language Models

Leveraging Argumentation Knowledge Graph for Interactive Argument Pair Identification.

Graph Chain-of-Thought: Augmenting Large Language Models by Reasoning on Graphs

Leveraging Structured Information for Explainable Multi-hop Question Answering and Reasoning

Exploring Discourse Structures for Argument Impact Classification

Pattern-Based Meta Graph Neural Networks for Argument Classifications