广东工业大学学报 ›› 2022, Vol. 39 ›› Issue (04): 24-31.doi: 10.12052/gdutxb.210062

• • 上一篇    下一篇

概率条件下基于双目标交替优化的知识表示模型

张欣, 王振友   

  1. 广东工业大学 数学与统计学院, 广东 广州 510520
  • 收稿日期:2021-04-25 出版日期:2022-07-10 发布日期:2022-06-29
  • 通信作者: 王振友(1979–)，男，教授，博士，主要研究方向为医学影像学、最优化理论及应用、数值计算等，E-mail：zywang@gdut.edu.cn
  • 作者简介:张欣(1998–)，女，硕士研究生，主要研究方向为数据分析、算法设计与分析
  • 基金资助:
    广东省基础与应用基础研究基金资助项目(2020B1515310001)

A Knowledge Representation Model Based on Bi-Objective Alternate Optimization Under Probability

Zhang Xin, Wang Zhen-you   

  1. School of Mathematics and Statistics, Guangdong University of Technology, Guangzhou 510520, China
  • Received:2021-04-25 Online:2022-07-10 Published:2022-06-29

HTML

PDF (PC)

2159

摘要： 针对TransD模型参数多和实体两种表示间没有关联的问题，提出一种改进的知识表示模型PTransD。通过减少实体投影数,并对实体进行聚类来减少参数量，同时利用K-L(Kullback-Leibler)散度限制实体投影和对应实体类，使其概率分布相同。在模型训练时，对三元组损失和K-L损失交替优化，从类间距大的实体类中替换实体，提高负例质量。最后，在知识图谱数据集上进行三元组分类和链接预测实验。结果表明，该模型的性能在各项指标上均有明显提高，可以应用于知识图谱的完善和推理等。

关键词: 知识图谱, 表示学习, 交替优化, 三元组分类, 链接预测

Abstract: Aiming at the problem that the TransD model has many parameters and the two representations of entities are not related, an improved knowledge representation model PTransD is proposed, which reduces the number of parameters by reducing the number of entity projections and clustering entities, while using K-L (Kullback-Leibler ) The divergence limits the entity projection to the same probability distribution as the corresponding entity class. During model training, the triple loss and K-L loss are alternately optimized, and the entities in the classes with large spacing between the entities are replaced to improve the quality of negative examples. Finally, based on the experimental results of triple classification and link prediction on the knowledge graph data set, the performance has been significantly improved in various indicators. It can be applied to the perfection and reasoning of knowledge map.

Key words: knowledge graph, representation learning, alternate optimization, triple classification, link prediction

中图分类号: 

  • TP3
[1] LENAT D B, PRAKASH M, SHEPHERD M. CYC: using common sense knowledge to overcome brittleness and knowledge acquisition bottlenecks [J]. AI Magazine, 1985, 6(4): 65.
[2] MILLER G A. WordNet [J]. Communications of the ACM, 1995, 38(11): 39-41.
[3] BOLLACKER K, COOK R, TUFTS P. Freebase: ashared database of structured general human knowledge[C]//Proceedings of the Twenty-Second AAAI Conference on Artificial Intelligence. Vancouver: AAAI Press, 2007: 1962-1963.
[4] JI S, PAN S, CAMBRIA E, et al. A survey on knowledge graphs: representation, acquisition and applications [EB/OL]. (2021-04-01)[2021-04-20]. https://arxiv.org/abs/2002.00388.
[5] GAO Y, LI Y, LIN Y, et al. Deep learning on knowledge graph for recommender system: a survey[EB/OL]. (2020-03-25) [2021-04-20]. https://arxiv.org/abs/2004.00387.
[6] HUANG X, ZHANG J, LI D, et al. Knowledge graph embedding based question answering [C]//Proceedings of the Twelfth ACM International Conference on Web Search and Data Mining. New York: ACM, 2019: 105-113.
[7] DAI Y, WANG S, XIONG N N, et al. A survey on knowledge graph embedding: approaches, applications and benchmarks [J]. Electronics, 2020, 9(5): 750.
[8] MIKOLOV T, SUTSKEVER I, CHEN K, et al. Distributed representations of words and phrases and their compositionality [J]. Advances in Neural Information Processing Systems, 2013, 26: 3111-3119.
[9] BORDES A, USUNIER N, GARCIA-DURAN A, et al. Translating embeddings for modeling multi-relational data[C]// Proceedings of the 26th International Conference on Neural Information Processing Systems. Lake Tahoe: NIPS, 2013: 2787-2795.
[10] LIN Y, LIU Z, SUN M, et al. Learning entity and relation embeddings for knowledge graph completion[C]//Proceedings of the AAAI Conference on Artificial Intelligence. Austin: AAAI Press, 2015: 2181-2187.
[11] JI G, HE S, XU L, et al. Knowledge graph embedding via dynamic mapping matrix[C] //Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing. Beijing: ACL, 2015: 687-696.
[12] WANG H, ZHANG F, ZHAO M, et al. Multi-task feature learning for knowledge graph enhanced recommendation[C]//Proceedings of the 2019 World Wide Web Conference. San Francisco: ACM, 2019: 2000-2010.
[13] WANG Z, ZHANG J, FENG J, et al. Knowledge graph embedding by translating on hyperplanes[C] //In Proceedings of the Twenty-eighth AAAI Conference on Artificial Intelligence. Québec City: AAAI Press, 2014: 1112–1119.
[14] DO K, TRAN T, VENKATESH S. Knowledge graph embedding with multiple relation projections[C]// 2018 24th International Conference on Pattern Recognition. Beijing: IEEE, 2018: 332-337.
[15] ZHU Q, ZHOU X, TAN J, et al. Learning knowledge graph embeddings via generalized hyperplanes[C] //Proceedings of the 18th International Conference on Computational Science. Wuxi: Springer, 2018: 624-638.
[16] ZEILER M D. Adadelta: an adaptive learning rate method[EB/OL]. (2012-12-22) [2021-04-20]. https://arxiv.org/abs/1212.5701v1.
[17] BORDES A, GLOROT X, WESTON J, et al. Joint learning of words and meaning representations for open-text semantic parsing[C]// Proceedings of the Fifteenth International Conference on Artificial Intelligence and Statistics. La Palma: PMLR, 2012: 127-135.
[18] BORDES A, WESTON J, COLLOBERT R, et al. Learning structured embeddings of knowledge bases[C]//In Proceedings of the Twenty-eighth AAAI Conference on Artificial Intelligence. San Francisco: AAAI Press, 2011: 301–306.
[19] BORDES A, GLOROT X, WESTON J, et al. A semantic matching energy function for learning with multi-relational data [J]. Machine Learning, 2014, 94(2): 233-259.
[1] 苏攀予, 黄健文. 基于知识图谱的BIM技术在工程造价领域的应用研究[J]. 广东工业大学学报, 2023, 40(02): 103-110,119.
[2] 汪培庄, 曾繁慧, 孙慧, 李兴森, 郭建威, 孟祥福, 何静. 知识图谱的拓展及其智能拓展库[J]. 广东工业大学学报, 2021, 38(04): 9-16.
[3] 林琳, 李小涛, 徐永利. 国内TRIZ理论及方法的研究热点及发展趋势[J]. 广东工业大学学报, 2021, 38(02): 10-19.
[4] 朱灿, 林豪慧, 向林芳. 新能源汽车领域研究进展及前沿动态：基于Citespace Ⅲ知识图谱分析[J]. 广东工业大学学报, 2020, 37(02): 45-52.
[5] 谢卫红, 李杰, 董策. 国内制造业转型升级研究热点与趋势——基于Citespace的知识图谱分析[J]. 广东工业大学学报, 2018, 35(06): 9-17.
[6] 刘贻新, 张光宇, 杨诗炜, 张玉磊. 国内战略生态位管理(SNM)研究的知识图谱:基于CiteSpaceⅢ的计量分析[J]. 广东工业大学学报, 2017, 34(03): 59-66.
[7] 孙莹, 冯冈平, 练思婷. 文献计量下品牌传播发展趋势分析及研究[J]. 广东工业大学学报, 2017, 34(02): 97-102.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
版权所有 © 2017 《广东工业大学学报》编辑部
地址：广州市东风东路729号广东工业大学 邮编：510090
电话：020-37626653,020-37626139,020-37626396
邮箱：xbzrb@gdut.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发