基于翻译模型的异质重边信息网络链路预测研究

doi:10.13232/j.cnki.jnju.2020.04.012

南京大学学报(自然科学版) ›› 2020, Vol. 56 ›› Issue (4): 541–548.doi: 10.13232/j.cnki.jnju.2020.04.012

基于翻译模型的异质重边信息网络链路预测研究

郑建兴^1,²(),李沁文¹,王素格^1,²,李德玉^1,²

^1.山西大学计算机与信息技术学院，太原，030006
^2.计算智能与中文信息处理教育部重点实验室（山西大学），太原，030006

收稿日期:2020-06-28 出版日期:2020-07-30 发布日期:2020-08-06
通讯作者: 郑建兴 E-mail:jxzheng@sxu.edu.cn
基金资助:
国家自然科学基金(61632011);山西省重点研发计划（国际科技合作）(201803D421024);山西省自然科学基金(201901D211174);山西省高等学校科技创新项目(2020L0001);山西省软科学研究一般项目(2018041015?3)

Research of link prediction based on translation model in heterogeneous multi⁃edge information network

Jianxing Zheng^1,²(),Qinwen Li¹,Suge Wang^1,²,Deyu Li^1,²

^1.School of Computer and Information Technology，Shanxi University，Taiyuan，030006，China
^2.Key Laboratory of Computational Intelligence and Chinese Information Processing of Ministry of Education （Shanxi University），Taiyuan，030006，China

Received:2020-06-28 Online:2020-07-30 Published:2020-08-06
Contact: Jianxing Zheng E-mail:jxzheng@sxu.edu.cn

摘要/Abstract

摘要：

在异质信息网络中，异质节点对象之间具有多元关系，形成异质重边信息网络.知识图谱表示旨在将实体和关系在低维的向量空间进行嵌入，可以用来学习异质重边信息网络中实体间的多元关系.首先通过注意力机制对异质重边信息网络中的多元关系进行融合表示，进而将异质节点的类型信息进行多元关系融合空间的映射，在多元关系融合空间上提出基于翻译的异质重边嵌入模型，用以学习异质节点之间的链路关系.最后，在MovieLens100k电影数据集上进行了异质节点多元关系的链路预测实验.实验结果表明，在异质重边信息网络中，基于改进的翻译模型在实体间链路预测性能方面要优于传统的知识表示方法，可以有效地提升链路预测的精度．

关键词: 异质重边信息网路, 链路预测, 翻译模型, 表示学习

Abstract:

In heterogeneous information network，heterogeneous nodes have multiple relations which can form heterogeneous multi?edge information network. Knowledge graph?based representation aims to embed object entities and relations into a low?dimensional vector space which can be used to learn the multiple relations between entities in heterogeneous multi?edge information network. In this paper，we first leverage fused representation of multiple relations for heterogeneous multi?edge information network in terms of attention mechanism. Then，projected matrices are adopted to map the types of heterogeneous nodes into fused spaces of multiple relations. More，in the fusion representation space of multiple relations，translation?based heterogeneous multi?edge embedding model is proposed to learn the link relations among heterogeneous nodes. Finally，link prediction experiments of heterogeneous multi?edge relations are carried out on MovieLens100k dataset. The experimental results demonstrate that the novel translation model is superior to traditional knowledge representation methods at the aspect of link prediction performance，which can effectively improve the accuracy of link prediction.

Key words: heterogeneous multi?edge information network, link prediction, translation model, representation learning

中图分类号:

TP391

郑建兴,李沁文,王素格,李德玉. 基于翻译模型的异质重边信息网络链路预测研究[J]. 南京大学学报(自然科学版), 2020, 56(4): 541–548.

Jianxing Zheng,Qinwen Li,Suge Wang,Deyu Li. Research of link prediction based on translation model in heterogeneous multi⁃edge information network[J]. Journal of Nanjing University(Natural Sciences), 2020, 56(4): 541–548.

图/表 8

图1

图2

表1

表2

表3

表4

TransHME在不同γ下的过滤结果"

$γ$	MRR	MR	Hits@5	Hits@3	Hits@1
4	0.6397	8.1205	0.8946	0.8223	0.4398
5	0.7373	6.0979	0.9157	0.8389	0.6054
6	0.7927	4.7605	0.9262	0.8795	0.6928
7	0.8190	7.0512	0.9503	0.8946	0.7304
8	0.8085	5.7771	0.9322	0.8886	0.7189

表4

表5

TransHME在不同λ下的过滤结果"

$λ$	MRR	MR	Hits@5	Hits@3	Hits@1
0.01	0.8028	4.0226	0.9367	0.8931	0.7108
0.1	0.8190	7.0512	0.9503	0.8946	0.7304
1.0	0.8063	5.7213	0.9292	0.8889	0.7139

表5

表6

TransHME在不同嵌入维度下的过滤结果"

$D$	MRR	MR	Hits@5	Hits@3	Hits@1
32	0.7610	7.4006	0.8931	0.8554	0.6611
64	0.7834	9.98071	0.9146	0.8650	0.6873
100	0.8190	7.0512	0.9503	0.8946	0.7304

表6

参考文献 28

1	李翠平，蓝梦微，邹本友等. 大数据与推荐系统. 大数据，2015，1(3)：23-35.
	Li C P，Lan M W，Zou B Y，et al. Big data and recommendation system. Big Data Research，2015，1(3)：23-35.
2	齐金山，梁循，李志宇等. 大规模复杂信息网络表示学习：概念、方法与挑战. 计算机学报，2018，41(10)：2394-2420.
	Qi J S，Liang X，Li Z Y，et al. Representation learning of large?scale complex information network：concepts，methods and challenges. Chinese Journal of Computers，2018，41(10)：2394-2420.
3	李宇琦，陈维政，闫宏飞等. 基于网络表示学习的个性化商品推荐. 计算机学报，2019，42(8)：1767-1778.
	Li Y Q，Chen W Z，Yan H F，et al. Learning graph?based embedding for personalized product re?commendation. Chinese Journal of Computers，2019，42(8)：1767-1778.
4	Zheng J X，Wang S G，Li D Y，et al. Personalized recommendation based on hierarchical interest overlapping community. Information Sciences，2019，479：55-75.
5	Shadbolt N,Hall W,Bernerslee T. The semantic web revisited. IEEE Intelligent Systems,2006,21(3):96-101.
6	王昊奋，漆桂林，陈华钧. 知识图谱：方法、实践与应用. 北京：电子工业出版社，2019：15-30.
7	Bordes A，Usunier N，Garcia?Durán A，et al. Translating embeddings for modeling multi?relational data∥Proceedings of the 26^th International Conference on Neural Information Processing Systems. Lake Tahoe，NV，USA：MIT Press，2013：2787-2795.
8	Ji G L，He S Z，Xu L H，et al. Knowledge graph embedding via dynamic mapping matrix∥Proceedings of the 53^rd Annual Meeting of the Association for Computational Linguistics and the 7^th International Joint Conference on Natural Language Processing. Beijing，China：Association for Computational Linguistics，2015：687-696.
9	Wang Z，Zhang J W，Feng J L，et al. Knowledge graph embedding by translating on hyperplanes∥Proceedings of 28^th AAAI Conference on Arti?cial Intelligence. Québec City，Canada：AAAI，2014：1112-1119.
10	Lin Y K，Liu Z Y，Sun M S，et al. Learning entity and relation embeddings for knowledge graph completion∥Proceedings of 29^th AAAI Conference on Arti?cial Intelligence. Austin，TX，USA：AAAI，2015：2181-2187.
11	Vaswani A，Shazeer N，Parmar N，et al. Attention is all you need∥Proceedings of the 31^st International Conference on Neural Information Processing System. Long Beach，CA，USA：NIPS，2017：1-12.
12	Goyal P，Ferrara E. Graph embedding techniques，applications and performance：a survey. Knowledge?Based Systems，2018，151：78-94.
13	Roweis S T，Saul L K. Nonlinear dimensionality reduction by locally linear embedding. Science，2000，290(5500)：2323-2326.
14	Luo D J，Huang H，Nie F P，et al. Cauchy graph embedding∥Proceedings of the 28^th International Conference on International Conference on Machine Learning. Bellevue，WA，USA：ACM，2011：553-560.
15	Perozzi B，Al?Rfou R，Skiena S. Deepwalk：online learning of social representations∥Proceedings of the 20^th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York，NY,USA：ACM，2014：701-710.
16	Tu C C，Zhang W C，Liu Z Y，et al. Max?margin deepwalk：discriminative learning of network representation∥Proceedings of the 25^th International Joint Conference on Artificial Intelligence. New York，NY，USA：Morgan Kaufmann，2016：3889-3895.
17	Grover A，Leskovec J. Node2vec：scalable feature learning for networks∥Proceedings of the 22^nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. San Francisco，CA,USA：ACM，2016：855-864.
18	Tang J，Qu M，Wang M Z，et al. Line：large?scale information network embedding∥Proceedings of the 24^th International Conference on World Wide Web. Florence，Italy：ACM，2015：1067-1077.
19	Wang D X，Cui P，Zhu W W. Structural deep network embedding∥Proceedings of the 22^nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. San Francisco，CA，USA：ACM，2016：1225-1234.
20	Dong Y X，Chawla N V，Swami A. Metapath2vec：scalable representation learning for heterogeneous networks∥Proceedings of the 23^rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. Nova Scotia，Canada：ACM，2017：135-144.
21	Wang X，Ji H Y，Shi C，et al. Heterogeneous graph attention network∥The World Wide Web Conference. San Francisco，CA，USA：ACM，2019：2022-2032.
22	Xiao H，Huang M L，Zhu X Y. TransG：A generative model for knowledge graph embedding∥Proceedings of the 54^th Annual Meeting of the Association for Computational Linguistics. Berlin，Germany：ACL，2016：2316-2325.
23	He S Z，Liu K，Ji G L，et al. Learning to represent knowledge graphs with gaussian embedding∥Proceedings of the 24^th ACM International on Conference on Information and Knowledge Management. Melbourne，Australia：ACM，2015：623-632.
24	Tu C C，Zhang Z Y，Liu Z Y，et al. Transnet：translation?based network representation learning for social relation extraction∥Proceedings of the 26^th International Joint Conference on Artificial Intelligence. Melbourne，Australia：Morgan Kaufmann，2017：2864-2870.
25	Wang X，Wang D X，Xu C N，et al. Explainable reasoning over knowledge graphs for recommendation∥Proceedings of the 33^rd AAAI Conference on Artificial Intelligence. Honolulu，HI，USA：AAAI，2019：5329-5336.
26	Wang H W，Zhang F Z，Wang J L，et al. RippleNet：Propagating user preferences on the knowledge graph for recommender systems∥Proceedings of the 27^th ACM International Conference on Information and Knowledge Management. Turin，Italy：ACM，2018：417-426.
27	Chen C，Zhang M，Ma W Z，et al. Ef?cient heterogeneous collaborative filtering without negative sampling for recommendation∥Proceedings of the 34^th AAAI Conference on Artificial Intelligence. New York，NY,USA：AAAI，2020：19-26
28	Sun Y Z，Han J W. Mining heterogeneous information networks：principles and methodologies. San Rafael：Morgan & Claypool Publishers，2012，1-159.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

Method	MRR	MR	Hits@5	Hits@3	Hits@1
TransE	0.0526	59.1370	0.0437	0.0136	0.0045
TransH	0.0613	55.5030	0.0738	0.0271	0.0030
TransR	0.0642	60.6901	0.0744	0.0372	0.0041
TransD	0.0711	49.8524	0.0723	0.0331	0.0075
TransE+Mule	0.7225	3.9593	0.9292	0.8524	0.5633
TransH+Mule	0.7132	7.3630	0.9057	0.8530	0.5627
TransR+Mule	0.7007	7.3991	0.9006	0.8298	0.5459
TransD+Mule	0.7071	10.3912	0.9029	0.8254	0.5431
TransHME	0.7511	7.4473	0.9322	0.8434	0.6250

Method	MRR	MR	Hits@5	Hits@3	Hits@1
TransE	0.1594	46.7319	0.2199	0.1521	0.0693
TransH	0.1617	44.2139	0.2349	0.1536	0.0693
TransR	0.1511	49.3733	0.2149	0.1501	0.0550
TransD	0.1791	39.1973	0.2304	0.1611	0.0889
TransE+Mule	0.7837	3.6235	0.9458	0.8961	0.6581
TransH+Mule	0.7707	6.9187	0.9252	0.8946	0.6416
TransR+Mule	0.7682	6.8976	0.9247	0.8675	0.6521
TransD+Mule	0.7663	9.8554	0.9230	0.8698	0.6409
TransHME	0.8190	7.0512	0.9503	0.8946	0.7304

[1]	钱付兰, 黄鑫, 赵姝, 张燕平. 基于路径相互关注的网络嵌入算法[J]. 南京大学学报(自然科学版), 2019, 55(4): 573-580.
[2]	齐小刚，张　权*. 认知无线电网络中一种基于主用户活动性预测的拓扑控制方法[J]. 南京大学学报(自然科学版), 2018, 54(4): 848-.

基于翻译模型的异质重边信息网络链路预测研究

Research of link prediction based on translation model in heterogeneous multi⁃edge information network

RichHTML

PDF (PC)

摘要/Abstract

引用本文

使用本文

图/表 8

参考文献 28

相关文章 2

Metrics

本文评价

推荐阅读 10