计算机科学

计算机科学 ›› 2023, Vol. 50 ›› Issue (12): 302-313.doi: 10.11896/jsjkx.221000170

• 人工智能 • 上一篇    下一篇

DL+:一种增强型双层知识图谱推理框架

武月佳, 周建涛   

  1. 内蒙古大学计算机学院(软件学院) 呼和浩特 010021
    生态大数据教育部工程研究中心 呼和浩特 010021
    蒙古文智能信息处理技术国家地方联合工程研究中心 呼和浩特 010021
    内蒙古自治区云计算与服务软件工程实验室 呼和浩特 010021
    内蒙古自治区社会计算与数据处理重点实验室 呼和浩特 010021
    内蒙古自治区纪检监察大数据重点实验室 呼和浩特 010021
    内蒙古自治区大数据分析技术工程实验室 呼和浩特 010021
  • 收稿日期:2022-10-23 修回日期:2023-03-06 出版日期:2023-12-15 发布日期:2023-12-07
  • 通讯作者: 周建涛(cszjtao@imu.edu.cn)
  • 作者简介:(wuyuejia@imudges.com)
  • 基金资助:
    国家自然科学基金(62162046);内蒙古科技攻关项目(2021GG0155);内蒙古自然科学基金重大项目(2019ZD15);内蒙古自然科学基金(2019GG372);内蒙古大学研究生创新创业专项经费(11200-5223737)

DL+:An Enhanced Double-layer Framework for Knowledge Graph Reasoning

WU Yuejia, ZHOU Jiantao   

  1. College of Computer Science(College of Software),Inner Mongolia University,Hohhot 010021,China
    Engineering Research Center of Ecological Big Data,Ministry of Education,Hohhot 010021,China
    National & Local Joint Engineering Research Center of Intelligent Information Processing Technology for Mongolian,Hohhot 010021,China
    Inner Mongolia Engineering Laboratory for Cloud Computing and Service Software,Hohhot 010021,China
    Inner Mongolia Key Laboratory of Social Computing and Data Processing,Hohhot 010021,China
    Inner Mongolia Key Laboratory of Discipline Inspection and Supervision Big Data,Hohhot 010021,China
    Inner Mongolia Engineering Laboratory for Big Data Analysis Technology,Hohhot 010021,China
  • Received:2022-10-23 Revised:2023-03-06 Online:2023-12-15 Published:2023-12-07
  • About author:WU Yuejia,born in 1996,Ph.D candidate,is a student member of China Computer Federation.Her main research interests include knowledge graph,knowledge graph representing and knowledge graph reasoning.
    ZHOU Jiantao,born in 1974,Ph.D,professor,Ph.D supervisor,is a member of China Computer Federation.Her main research interests include cloud computing technology,social network recommendation technology,software engineering and so on.
  • Supported by:
    National Natural Science Foundation of China(62162046),Inner Mongolia Science and Technology Project(2021GG0155),Natural Science Foundation of Major Research Plan of Inner Mongolia(2019ZD15),Natural Science Foundation of Inner Mongolia,China(2019GG372) and Special Fund for Graduate Innovation and Entrepreneurship of Inner Mongolia University(11200-5223737).

PDF (PC)

2315

摘要： 知识图谱是图数据库的一个重要研究领域,它可以形式化地描述现实世界中的事物及其关系,但其不完整性和稀疏性阻碍了其在诸多领域中的应用。知识图谱推理技术旨在根据知识图谱中已有的知识来推断新的知识或识别错误的知识以完善知识图谱。尽管现有的各类推理方法可以获得部分有效知识,但仍然存在获取路径不全、忽略局部信息和引入噪声等问题。基于此,发现且明确提出路径连通性差问题并证明推理有效性与实体间路径连通比率呈正相关规律,进一步提出一种用于增强现有推理方法性能的双层框架DL+。模型第一层是知识增广器,主要利用社区发现算法在初始知识图谱上提取实体邻域信息,构建新知识以增广知识规模,然后设计社区剪枝优化去除构建时引入的噪声,最后将增广后的知识图谱抽取还原为与初始图谱表示相同的结构并输出到第二层以保证模型“即插即用”的特性。第二层是知识推理机,通过在知识增广后的图谱上进行学习推理以达到增强现有知识图谱推理模型的目的,使模型可以在图谱路径连通性比率较高的情况下获得更优的推理结果。最终在4个标准知识图谱数据集上进行的大量实验结果表明DL+算法可以有效缓解实体间路径连通性差的问题,与9类基准方法相比,所提算法的预测精度平均提高了4.798%。

关键词: 知识图谱, 知识图谱推理, 社区发现, 路径连通性, 链接预测

Abstract: As an important research field of the graph database,the knowledge graph(KG) can formally describe things and their relationships in the real world.However,its incompleteness and sparsity hinder its application in many fields.The knowledge graph reasoning(KGR) technology aims to complete the knowledge graph by inferring new knowledge or identifying wrong knowledge according to the existing knowledge in the knowledge graph.Although existing reasoning methods can obtain partially effective knowledge paths,there are still some problems such as incomplete acquisition paths,ignoring local information,introducing noise.Based on this,this paper finds and explicitly proposes the problem of poor path connectivity,proves that the reasoning validity is positively correlated with the path connectivity ratio between entities,and further proposes a double-layer framework DL+ which is used to enhance the performance of existing reasoning methods.The first layer is a knowledge augmenter,which mainly uses the community discovery algorithm to extract the entity neighborhood information on the initial KG and build new knowledge to expand the knowledge scale,and then designs a community pruning optimization method to remove the noise introduced in the construction.Finally,the augmented KG is extracted and restored to the same structure as the initial KG representation and output to the second layer to ensure the “plug-and-play” feature of the model.The second layer is a knowledge reasoner,which can enhance the existing KGR model by learning and reasoning on the KG after knowledge augmentation,so that the model can obtain better reasoning results when the graph path connectivity ratio is high.Finally,a large number of experimental results on four standard KG datasets show that the DL+ can effectively alleviate the problem of poor path connectivity between entities,and improve the average prediction accuracy by 4.798% compare with nine types of benchmark methods.

Key words: Knowledge graph, Knowledge graph reasoning, Community discovery, Path connectivity, Link prediction

中图分类号: 

  • TP391
[1]BOLLACKER K D,EVANS C,PARITOSG P,et al.Freebase:a collaboratively created graph database for structuring human knowledge[C]//Proceedings of the ACM SIGMOD Interna-tional Conference on Management of Data.New York:ACM,2008.
[2]LEHMANN J.DBpedia:A Nucleus for a Web of Open Data[C]//Proceedings of the Semantic Web,International Semantic Web Conference,Asian Semantic Web Conference.Busan,Korea,2007.
[3]MILLER G A.WordNet:a lexical database for English[J].Communications of the ACM,1995,38(11):39-41.
[4]SUN Y W,CHENG G,LI X,et al.Complex Question Answe-ring Method for Explainable Knowledge Graph Based on Graph Matching Network[J].Journal of Computer Research and Development,201,58(12):2673-2683.
[5]LU L,KONG F.Dialogue-oriented Entity Relation Extractionwith Integrated Knowledge[J].Computer Science,2022,49(5):200-205.
[6]GE F B,SHEN X.Application research of knowledg-e graph technology in patent semantic retrieval[J].China Inventions and Patents,2022,19(1):10-18.
[7]GUAN S P,JIN X L,JIA Y T,et al.Research progress inKnowledge Reasoning based on Knowledge Graph[J].Journal of Software,2018,29(10):2966-2994.
[8]DETTMERS T,MINERVINI P,STENETORP P,et al.Convolutional 2D Knowledge Graph Embeddings[C]//Proceddings of the 32nd AAAI Conference on Artificial Intelligence(AAAI-18).New Orleans,LA,USA.2017.
[9]ZHU G,IGLESIAS C A.Exploiting semantic similarity fornamed entity disambiguation in knowledge graphs[J].Expert Systems with Application,2018(101):8-24.
[10]WANG X,HE X,CAO Y,et al.KGAT:Knowledge Graph Attention Network for Recommendation[C]//Procedding of the Knowledge Discovery and Data Mining.Anchorage:ACM,2019.
[11]PUJARA J,AUGUSTINE E,GETOOR L.Sparsity and Noise:Where Knowledge Graph Embeddings Fall Short[C]//Procee-dings of the 2017 Conference on Empirical Methods in Natural Language Processing.Copenhagen,2017.
[12]LV X,HAN X,HOU L,et al.Dynamic Anticipation and Completion for Multi-Hop Reasoning over Sparse Knowledge Graph[C]//Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing.Punta Cana,2020.
[13]LAO N,WILLIAM W C.Relational retrieval using a combination of path-constrained random walks[J].Machine Learning,2010,81:53-67.
[14]SHI B,WENINGER T.ProjE:Embedding Projection for Know-ledge Graph Completion[C]//Proceedings of the AAAI Confe-rence on Artificial Intelligence.Phoenix,2016.
[15]SEN P,CARVALHO B,ABDELAZIZ I,et al.Combining Rules and Embeddings via Neuro-Symbolic AI for Knowledge Base Completion[J].arXiv:2109.09566,2021.
[16]ZHANG Z,ZHUANG F,ZHU H,et al.Relational Graph Neural Network with Hierarchical Attention for Knowledge Graph Completion[C]//Proceedings of the AAAI Conference on Artificial Intelligence.2020.
[17]NEWMAN M.Modularity and community structure in networks[C]//Proceedings of the National Academy of Sciences of the United States of America.2006.
[18]BLONDEL V D,GUILLAUME J L,LAMBIOTTE R,et al.Fast unfolding of communities in large networks[J].arXiv:0803.0476,2008.
[19]GARDNER M,MITCHELL T.Efficient and Expressive Know-ledge Base Completion Using Subgraph Feature Extraction[C]//Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing.2015.
[20]XIONG W,HOANG T,WANG W Y.DeepPath:A Reinforcement Learning Method for Knowledge Graph Reasoning[C]//Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing.Copenhagen,2017.
[21]SCHLICHTKRULL M,KIPG T N,BLEOM P,et al.Modeling Relational Data with Graph Convolutional Networks[C]//Proceedings of the Extended Semantic Web Conference.Heraklion,2018.
[22]YAO L,MAO C,LUO Y.KG-BERT:BERT for KnowledgeGraphCompletion[C]//Proceedings of the AAAI Conference on Artificial Intelligence.2020.
[23]WANG S,ZHONG Y,WANG C.Attention Relational GraphConvolution Networks for Relation Prediction in Knowledge Graphs[C]//2021 4th International Conference on Advanced Algorithms and Control Engineering(ICAACE 2021).2021.
[24]WANG L,ZHAO W,WEI Z,et al.SimKGC:Simple Contrastive Knowledge Graph Completion with Pre-trained Language Mo-dels[C]//Proceedings of the Association for Computational Linguistics.2022.
[25]CUI W Y,CHEN X R.Instance-based Learning for Knowledge Base Completion[C]//Proceedings of the Neural Information Processing Systems.2022.
[26]TANG Z,PEI S,ZHANG Z,et al.Positive-Unlabeled Learning with Adversarial Data Augmentation for Knowledge Graph Completion[C]//Proceedings ofthe International Joint Confe-rence on Artificial Intelligence.2022.
[27]QUAN W,JING L,LUO Y,et al.Knowledge Base Completion via Coupled Path Ranking[C]//Proceedings of the Association for Computational Linguistics.2016:1308-1318.
[28]QIAO L,JIANG L,HAN M,et al.Hierarchical Random WalkInference in Knowledge Graphs[C]//Proceedings of the 39th International ACM SIGIR conference on Research and Development in Information Retrieval.Pisa:ACM,2016:445-454.
[29]JIA Y,WANG Y,LIN H,et al.Locally adaptive translation for knowledge graph embedding[C]//Proceedings of the AAAI Conference on Artificial Intelligence.2016:992-998.
[30]NICKEL M,TRESP V,KRIEGEL HP.A Three-Way Model for Collective Learning on Multi-Relational Data[C]//Proceedings of the International Conference on Machine Learning.2011:809-816.
[31]HAN X,HUANG M,YU H,et al.TransG:A Generative Mixture Model for Knowledge Graph Embedding[J].arXiv:1509.05488,2015.
[32]LI R,CHENG X.DIVINE:A Generative Adversarial ImitationLearning Framework for Knowledge Graph Reasoning[C]//Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing.2019.
[33]TIAN A,ZHANG C,MIAO R,et al.RA-GCN:Relational Ag-gregation Graph Convolutional Network for Knowledge Graph Completion[C]//Proceedings of the 12th InternationalConfe-rence on Machine Learning and Computing.2020:580-586.
[34]VASHISHTH S,SANYAL S,NITIN V,et al.Composition-based Multi-Relational Graph Convolutional Networks[C]//Proceedings of the International Conference on Learning Representations.2020.
[35]WANG H,LIN H Z,LU L Y.Knowledge graph reasoning algorithm based on Att_GCN model[J].Computer Engineering and Applications,2020,56(9):183-189.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发