Abstract:Knowledge graph is of great research value to artificial intelligence,which has been extensively applied in the fields of semantic search and question answering,etc.Knowledge graph representation transforms a large-scale knowledge graph comprising entities and relations into a continuous vector space.To this end,there have been a number of models and methods proposed for knowledge embedding.Among them,TransE is a classic translation-based method that is of low model complexity,high computational efficiency,as well as good capability of expressing knowledge.However,TransE still has two flaws:one is that it utilizes inflexible Euclidean distance as metric,and treats each feature dimension identically,hence,the model accuracy may be interfered by irrelevant dimensions;the other is that it has limitations in dealing with complex relations including reflexive,one-to-many,many-to-one and many-to-many relations.Currently,there has not been a single method that resolves the flaws simultaneously,and thus,we propose a revised translation-based method for knowledge graph representation,namely,TransAH.For the first flaw,TransAH adopts an adaptive metric,replacing Euclidean distance with weighted Euclidean distance by adding a diagonal weight matrix,which assigns different weights to every feature dimension.As to the second,inspired by TransH,it introduces the relation-oriented hyperspace model,projecting head and tail entities to hyperspace of a given relation for distinction.At last,empirical studies on public real knowledge graph datasets analyze and verify the effectiveness of the proposed method.Comprehensive comparative experiments using two tasks-link prediction and triplet classification show that,in contrast to the existing models and methods,TransAH achieves remarkable improvement in various aspects and demonstrates its superiority.

Knowledge Representation Learning Via Dynamic Relation Spaces

Knowledge Graph Embedding by Dynamic Translation

Transr*: Representation Learning Model By Flexible Translation And Relation Matrix Projection

GTrans: Generic Knowledge Graph Embedding Via Multi-State Entities and Dynamic Relation Spaces

Incorporating Domain and Range of Relations for Knowledge Graph Completion.

A Knowledge Representation Method for Multiple Pattern Embeddings Based on Entity-Relation Mapping Matrix

Knowledge Graph Embedding with Diversity of Structures

TransRPRT: A Knowledge Representation Learning Model Based on Relationship Transfer and Relationship Mapping Properties

Modeling Relation Path for Knowledge Graph Via Dynamic Projection.

TransH-RA: A Learning Model of Knowledge Representation by Hyperplane Projection and Relational Attributes

TransE-MTP: A New Representation Learning Method for Knowledge Graph Embedding with Multi-Translation Principles and TransE

Text-Enhanced Representation Learning for Knowledge Graph.

TransP: A New Knowledge Graph Embedding Model by Translating on Positions*

TripleRE: Knowledge Graph Embeddings via Tripled Relation Vectors

Knowledge Representation Learning Method Based on Semantic Enhancement of External Information

Learning Entity and Relation Embeddings for Knowledge Graph Completion

Distributed Representation Learning for Knowledge Graphs with Entity Descriptions

Learning Attention-Based Translational Knowledge Graph Embedding via Nonlinear Dynamic Mapping

TransMS: Knowledge Graph Embedding for Complex Relations by Multidirectional Semantics

A Revised Translation-Based Method for Knowledge Graph Representation