计算机科学 ›› 2024, Vol. 51 ›› Issue (6A): 230600209-8.doi: 10.11896/jsjkx.230600209
朱玉亮, 刘俊涛, 饶子昀, 张毅, 曹万华
ZHU Yuliang, LIU Juntao, RAO Ziyun, ZHANG Yi, CAO Wanhua
摘要： 知识推理技术是解决知识图谱缺失问题所提出的方法,并在近年来不断发展。为了解决推理中准确度低、可解释性差、适用性不强等问题,提出了一种融合注意力机制和HousE的知识推理模型Att-HousE。该模型由一个带注意力机制的规则生成器和一个带HousE嵌入的规则预测器组成,规则生成器生成推理需要的规则并传入预测器,预测器更新并得到不同规则的得分,然后通过EM算法不断训练优化生成器与预测器。具体而言,该模型是建立在RNNLogic的基础上并作出改进,注意力机制可以选取更值得关注的关系作为规则,提高了模型准确度,HousE嵌入则在处理复杂关系上更具有灵活性,并适用于建立多边关系。在公开实验数据集上的结果表明,Att-HousE在FB15K-237上做推理任务时,MRR指标整体比RNNLogic高出6.3%;在稀疏数据集WN18RR上,Hits@10指标整体比RNNLogic高出2.7%,证明了引入HousE和注意力机制后可以更全面地抓取和形成多边关系,提升知识推理的精度。
中图分类号:
| [1]HUANG Q H,YU J,LIAO X,et al.Review of KnowledgeGraph Research[J].Computer Systems & Applications,2019,28(6):1-12. [2]BOLLACKER K,EVANS C,PARITOSH P,et al.Freebase:A collaboratively created graph database for structuring human knowledge[C]//Proceedings of 2008 ACM SIGMOD International Conference on Management of Data.2008:1247-1250. [3]MILLERG A.WordNet:A lexical database for English[J].Communications of the ACM,1995,38(11):39-41. [4]CARLSON A,BETTERIDGE J,KISIEL B,et al.Toward an ar-chitecture for never-ending language learning[C]//Proceedings of the 24th AAAI Conference on Artificial Intelligence.Menlo Park:AAAI,2010:1306-1313. [5]XU Z L,SHENG Y P,HE L R,et al.Overview of Knowledge Graph Technology[J].Journal of University of Electronic Science and Technology of China,2016,45(4):589-606. [6]TIAN L,ZHANG J C,ZHANG J H,et al.Overview of Know-ledge graph:Representation,Construction,Reasoning and Knowledge Hypergraph Theory[J].Computer Systems & Applications,2021,41(8):2161-2186. [7]LAO N,COHEN W W.Relational retrieval using a combination of path-constrained random walks[J].Machine Learning,2010,81(1):53-67. [8]SUN S,CHEN J,LIU D,et al.A Posterior-Based Method forMarkov Logic Networks Parameters Learning[C]//2006 5th IEEE International Conference on Cognitive Informatics.Beijing,China,2006:529-534. [9]YANG F,YANG Z L,COHEN W W.Differentiable Learning of Logical Rules for Knowledge Base Reasoning[J].arXiv:1702.08367,2017. [10]ASCHEL T R,RIEDEL S.End-to-end differentiable proving[J].Advances in Neural Information Processing Systems,2017,2017(December):3789-3801. [11]BORDES A,USUNIER N,GARCIA-DURAN A,et al.Translating embeddings for modeling multi-relational data[C]//Advances in Neural Information Processing Systems.2013:2787-2795. [12]SUN Z,DENG Z H,NIE J Y,et al.Rotate:Knowledge graph embedding by relational rotation in complex space[C]//7th International Conference on Learning Representations.2019. [13]LI R et al.HousE:Knowledge Graph Embedding with Householder Parameterization[J].International Conference on Machine Learning,2022. [14]KHOT T,NATARAJAN S,KERSTING K,et al.Learningmarkov logic networks via functional gradient boosting[C]//ICDM.2011. [15]MEILICKE C,CHEKOL M W,RUFFINELLI D,et al.Anytime bottom-up rule learning for knowledge graph completion[C]//IJCAI,2019. [16]MINERVINI P,BOŠNJAK M,ROCKTÄSCHELT,et al.Dif-ferentiable Reasoning on Large Knowledge Bases and Natural Language[J].arXiv:1912.0824,2019. [17]RONAN C,WESTON J.A unified architecture for natural language processing:deep neural networks with multitask learning[C]//International Conference on Machine Learning.2008. [18]TROUILLON T,WELBL J,RIEDEL S,et al.Complex embed-dings for simple link prediction[C]//ICML.2016. [19]ZHANG S,TAY Y,YAO L N,et al.Quaternion knowledgegraph embeddings[C]//Proceedings of the 33rd Conference on Neural Information Processing Systems.2019:2735. [20]XIONG W,HOANG T,WANG W Y.Deeppath:A reinforcement learning method for knowledge graph reasoning[J].arXiv:1707.06690,2017. [21]LIN Y,LIU Z,LUAN H,et al.Modeling relation paths for representation learning of knowledge bases[C]//Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing.2015:705-714. [22]LI N,SHEN Q,SONG R,et al.MEduKG:A Deep-Learning-Based Approach for Multi-Modal Educational Knowledge Graph Construction[J].Information,2022,13(2):91. [23]QU M,CHEN J,XHONNEUX L,et al.RNNLogic:Learning Logic Rules for Reasoning on Knowledge Graphs[J].arXiv:2010.04029,2020. [24]TOUTANOVA K,CHEN D Q.Observed versus latent features for knowledge base and text inference[C]//Workshop on Continuous Vector Space Models and their Compositionality.2015. [25]DETTMERS T,MINERVINI P,STENETORP P,et al.Convolutional 2d knowledge graph embeddings[C]//AAAI.2018. |
|
||
首页