Knowledge Transfer for Out-of-Knowledge-Base Entities: A Graph Neural Network Approach

Abstract

Knowledge base completion (KBC) aims to predict missing information in a knowledge base.In this paper, we address the out-of-knowledge-base (OOKB) entity problem in KBC:how to answer queries concerning test entities not observed at training time. Existing embedding-based KBC models assume that all test entities are available at training time, making it unclear how to obtain embeddings for new entities without costly retraining. To solve the OOKB entity problem without retraining, we use graph neural networks (Graph-NNs) to compute the embeddings of OOKB entities, exploiting the limited auxiliary knowledge provided at test time.The experimental results show the effectiveness of our proposed model in the OOKB setting.Additionally, in the standard KBC setting in which OOKB entities are not involved, our model achieves state-of-the-art performance on the WordNet dataset. The code and dataset are available at https://github.com/takuo-h/GNN-for-OOKB

This paper has been accepted by IJCAI17

References in corpus (4)

• Adam: A Method for Stochastic Optimization
• A Review of Relational Machine Learning for Knowledge Graphs
• Traversing Knowledge Graphs in Vector Space
• Neighborhood Mixture Model for Knowledge Base Completion

Cited by in corpus (34)

• Graph Neural Networks: A Review of Methods and Applications
• Graph Learning: A Survey
• Inductive Relation Prediction by Subgraph Reasoning
• Self-supervised Consensus Representation Learning for Attributed Graph
• Precise Learning of Source Code Contextual Semantics via Hierarchical Dependence Structure and Graph Attention Networks
• AGL: a Scalable System for Industrial-purpose Graph Machine Learning
• Knowledge Infused Learning (K-IL): Towards Deep Incorporation of Knowledge in Deep Learning
• Multi-Stage Self-Supervised Learning for Graph Convolutional Networks on Graphs with Few Labels
• Say No to the Discrimination: Learning Fair Graph Neural Networks with Limited Sensitive Attribute Information
• NodePiece: Compositional and Parameter-Efficient Representations of Large Knowledge Graphs
• Text Level Graph Neural Network for Text Classification
• A Tailored Convolutional Neural Network for Nonlinear Manifold Learning of Computational Physics Data using Unstructured Spatial Discretizations
• Logic Attention Based Neighborhood Aggregation for Inductive Knowledge Graph Embedding
• Coherent Comment Generation for Chinese Articles with a Graph-to-Sequence Model
• EnGN: A High-Throughput and Energy-Efficient Accelerator for Large Graph Neural Networks
• Improving Coreference Resolution by Leveraging Entity-Centric Features with Graph Neural Networks and Second-order Inference
• Graph-Aware Transformer: Is Attention All Graphs Need?
• Out-of-Sample Representation Learning for Multi-Relational Graphs
• Graph Neural Processes: Towards Bayesian Graph Neural Networks
• Continuous Graph Flow
• HR-RCNN: Hierarchical Relational Reasoning for Object Detection
• Explicit Pairwise Factorized Graph Neural Network for Semi-Supervised Node Classification
• Graph Convolutional Networks in Feature Space for Image Deblurring and Super-resolution
• Graph Convolution Networks for Probabilistic Modeling of Driving Acceleration
• Neural Consciousness Flow
• Light Field Saliency Detection with Dual Local Graph Learning andReciprocative Guidance
• OntoZSL: Ontology-enhanced Zero-shot Learning
• Convolutions for Spatial Interaction Modeling
• Adversarial Stein Training for Graph Energy Models
• Fractional order graph neural network
• Can NetGAN be improved on short random walks?
• Leveraging Semantics for Incremental Learning in Multi-Relational Embeddings
• Dense Graph Convolutional Neural Networks on 3D Meshes for 3D Object Segmentation and Classification
• Meta Graph Attention on Heterogeneous Graph with Node-Edge Co-evolution