知识图谱是指知识库中存储大量的三元组，通过三元组（头实体、关系实体、尾实体）描述物质之间的联系，从而提供大量真实和有价值的信息。在一般情况下，由于信息的不完全性，它的建立只能依靠全部事实中的一小部分，而大量的实体间隐性关联却没有得到充分利用。可以利用预测中缺失的部分信息进行处理。所有的知识图谱，都需要不断地完善，甚至推论出新的知识。基于此，本文提出一种基于神经张量网络的改进 TuckER分解算法。
Knowledge map usually refers to a knowledge base that stores a large number of triples. It uses triples (head entity, relationship, tail entity) to represent the relationship between entities, thus providing practical and valuable information. Generally, they are incomplete, because their construction may be based on only a small part of all facts, and the implicit relationship between a large number of entities has not been fully utilized. We can solve this problem by predicting the missing parts.Any existing knowledge map needs to constantly improve knowledge itself, or even infer new knowledge through completion. In this paper, a new link prediction algorithm, called improved TuckER NTN (ImTu NTN), is proposed by combining the improved TuckER decomposition algorithm and neural tensor network (NTN), which can effectively improve the accuracy of knowledge map link prediction.
2023,45(24): 166-170 收稿日期：2022-12-23
DOI：10.3404/j.issn.1672-7649.2023.24.030
分类号：U664
作者简介：李壮(1983-),男,博士,高级工程师,研究方向为系统顶层设计
参考文献：
[1] 陈恒,祁瑞华,朱毅,等.球坐标建模语义分层的知识图谱补全方法[J]. 计算机工程与应用，2020,57(15):101-108.
CHEN Heng, QI Ruihua, ZHU Yi, et al. A knowledge graph completion method for semantic layering in spherical coordinate modeling [J] Computer Engineering and Applications, 2020, 57 (15): 101-108
[2] SUCHANEK F M, et al. Yago: a core of semantic knowledge[C] //Proceedings of the 16th International World Wide Web Conference, 2007: 697-706.
[3] BOLLACKER K D, et al. Freebase: a collaboratively created graph database for structuring human knowledge[C] //SIGMOD Conference, 2008: 1247-1250.
[4] LEHMANN JD B. A Nucleus for a Web of Open Data[C]//Semantic Web, International Semantic Web Conference, Asian Semantic Web Conference, Iswc + Aswc, Busan, Korea, November, 2007.
[5] CARLSON A, BETTERIDGE J, KISIEL B, et al. Toward an architecture for never-ending language learning[C]// Proceedings of the AAAI Conference on Artificial Intelligence, 2010, 24(1), 1306-1313.
[6] YANG B, MITCHELL T. Leveraging knowledge bases in LSTMs for improving machine reading [C]//Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics, 2017.
[7] DONG X, GABRILOVICH E, HEITZ G, et al. Knowledge vault:A web-scale approach to probabilistic knowledge fusion[C]//Proceedings of the 20th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2014: 601-610.
[8] 陈恒, 李冠宇, 祁瑞华, 等. 改进的Tucker分解知识图谱补全算法[J]. 数学的实践与认识，2020, 50 (16) ：164-176.
[9] NICKEL M, TRESP V, KRIEGEL H P.A Three-way model for collective learning on multi-relational data[C]//ICML, 2011, 11: 809-816.
[10] YANG B, YIH W, HE X, et al. Embedding entities and relations for learning and inference in knowledge bases[J].arXiv preprint arXiv:1412.6575,2014.
[11] Théo Trouillon, WELBL J, RIEDEL S, et al. Complex embeddings for simple link prediction[C]//ICML，2016：2071-2080.
[12] KAZEMI SEYED M, DAVID L. Poole. SimplE embedding for link prediction in knowledge graphs[J]. NeurIPS, 2018.
[13] YANG FAN, et al. Differentiable learning of logical rules for knowledge base reasoning[J]. NIPS, 2017.
[14] HITCHCOCK FRANK LAUREN. The expression of a tensor or a polyadic as a sum of products[J]. Journal of Mathematics and Physics 6: 164-189.
[15] SCHILICHTKRULL M, et al. Modeling relational data with graph convolutional networks[J]. ArXiv abs/1703.06103, 2018.
[16] DAS RAJARSNI, et al. Go for a walk and arrive at the answer: reasoning over paths in knowledge bases using reinforcement learning[J]. ArXiv abs/1711.05851, 2018.
[17] DETTMERS TIM, et al. Convolutional 2D knowledge graph embeddings[C]//AAAI , 2018.
[18] TUCKER, LEDYARD R. Some mathematical notes on three-mode factor analysis[J]. Psychometrika, 1966, 31(3): 279-311.
[19] BSLAZEVIC IVANA, et al. TuckER: Tensor Factorization for Knowledge Graph Completion[J]. EMNLP , 2019.
[20] SOCHER RICHARD, et al. Reasoning with neural tensor networks for knowledge base completion[J]. NIPS, 2013.
