Abstract:Semantic networks, such as the knowledge graph, can represent the knowledge leveraging the graph structure. Although the knowledge graph shows promising values in natural language processing, it suffers from incompleteness. This paper focuses on knowledge graph completion by predicting linkage between entities, which is a fundamental yet critical task. Semantic matching is a potential solution as it can deal with unseen entities, which the translational distance based methods struggle with. However, to achieve competitive performance as translational distance based methods, semantic matching based methods require large-scale datasets for the training purpose, which are typically unavailable in practical settings. Therefore, we employ the language model and introduce a novel knowledge graph architecture named LP-BERT, which contains two main stages: multi-task pre-training and knowledge graph fine-tuning. In the pre-training phase, three tasks are taken to drive the model to learn the relationship from triples by predicting either entities or relations. While in the fine-tuning phase, inspired by contrastive learning, we design a triple-style negative sampling in a batch, which greatly increases the proportion of negative sampling while keeping the training time almost unchanged. Furthermore, we propose a new data augmentation method utilizing the inverse relationship of triples to improve the performance and robustness of the model. To demonstrate the effectiveness of our method, we conduct extensive experiments on three widely-used datasets, WN18RR, FB15k-237, and UMLS. The experimental results demonstrate the superiority of our methods, and our approach achieves state-of-the-art results on WN18RR and FB15k-237 datasets. Significantly, Hits@10 indicator is improved by 5% from previous state-of-the-art result on the WN18RR dataset while reaching 100% on the UMLS dataset.

Knowledge Graph Completion Via Multi-feature Learning

Knowledge Representation Learning with Entity Descriptions, Hierarchical Types, and Textual Relations

Calculation of melatonin and resveratrol effects on steatosis hepatis using soft computing methods

KG-BERT: BERT for Knowledge Graph Completion

Multi-task Pre-training Language Model for Semantic Network Completion

Multi-level Shared Knowledge Guided Learning for Knowledge Graph Completion

Multi-hop neighbor fusion enhanced hierarchical transformer for multi-modal knowledge graph completion

Federated Knowledge Graph Completion Via Embedding-Contrastive Learning

Multi-concept Representation Learning for Knowledge Graph Completion

Multi-perspective Improvement of Knowledge Graph Completion with Large Language Models

Multi-perspective knowledge graph completion with global and interaction features

Knowledge Graphs Meet Multi-Modal Learning: A Comprehensive Survey

Multi-relational knowledge graph completion method with local information fusion

Exploring Large Language Models for Knowledge Graph Completion

MEGA: Meta-Graph Augmented Pre-Training Model for Knowledge Graph Completion

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

Hybrid Transformer with Multi-level Fusion for Multimodal Knowledge Graph Completion

Knowledge Graph Completion Method of Combining Structural Information with Semantic Information

Towards Robust Knowledge Graph Embedding via Multi-task Reinforcement Learning

MSGE: A Multi-step Gated Model for Knowledge Graph Completion

Deep Sparse Latent Feature Models for Knowledge Graph Completion