多网络混合嵌入学习的服务推荐方法研究

doi:10.3778/j.issn.1673-9418.2101032

计算机科学与探索 ›› 2022, Vol. 16 ›› Issue (7): 1529-1542.DOI: 10.3778/j.issn.1673-9418.2101032

多网络混合嵌入学习的服务推荐方法研究

王雪纯¹, 吕晟凯¹, 吴浩², 何鹏¹^,³^,⁺(), 曾诚¹

1.湖北大学计算机与信息工程学院,武汉 430062
2.华中科技大学计算机科学与技术学院,武汉 430074
3.湖北大学数学与统计学学院应用数学湖北省重点实验室,武汉 430062

收稿日期:2021-01-07 修回日期:2021-03-04 出版日期:2022-07-01 发布日期:2021-03-12
作者简介:王雪纯（1996—）,女,湖北十堰人,硕士研究生,主要研究方向为表示学习、神经网络、服务计算。
WANG Xuechun, born in 1996, M.S. candidate. Her research interests include representation learning, neural networks and service computing.
吕晟凯（2001—）,男,湖北襄阳人,主要研究方向为深度学习、个性化推荐系统。
LYU Shengkai, born in 2001. His research interests include deep learning and personalized recommendation system.
吴浩（1998—）,男,湖北黄冈人,硕士研究生,主要研究方向为表示学习、神经网络、服务推荐。
WU Hao, born in 1998, M.S. candidate. His research interests include representation learning, neural networks and service recommendation.
何鹏（1988—）,男,江西宜春人,博士,副教授,主要研究方向为面向服务的软件工程、软件质量分析、缺陷预测。
HE Peng, born in 1988, Ph.D., associate professor. His research interests include service-oriented software engineering, software quality analysis and defect prediction.
曾诚（1976—）,男,湖北武汉人,博士,教授,主要研究方向为服务计算、机器学习、软件工程。
ZENG Cheng, born in 1976, Ph.D., professor. His research interests include services computing, machine learning and software engineering.
基金资助:
国家重点研发计划(2018YFB1003801);国家自然科学基金(61832014);国家自然科学基金(61902114);应用数学湖北省重点实验室开放基金(HBAM201901)

Research on Service Recommendation Method of Multi-network Hybrid Embed-ding Learning

WANG Xuechun¹, LYU Shengkai¹, WU Hao², HE Peng¹^,³^,⁺(), ZENG Cheng¹

1. School of Computer and Information Engineering, Hubei University, Wuhan 430062, China
2. School of Computer Science and Technology, Huazhong University of Science and Technology, Wuhan 430074, China
3. Hubei Key Laboratory of Applied Mathematics, Faculty of Mathematics and Statistics, Hubei University, Wuhan 430062, China

Received:2021-01-07 Revised:2021-03-04 Online:2022-07-01 Published:2021-03-12
Supported by:
the National Key Research and Development Program of China(2018YFB1003801);the National Natural Science Foundation of China(61832014);the National Natural Science Foundation of China(61902114);the Open Foundation of Hubei Key Laboratory of Applied Mathematics(HBAM201901)

摘要/Abstract

摘要：

网络嵌入是将网络节点投影到一个向量空间,从而有效地提取网络中各节点的特征信息。在服务推荐领域,已有研究表明引入网络嵌入方法能有效缓解推荐过程中数据稀疏等问题。但现有的网络嵌入方法多针对某一种特定结构的网络,并没有从根源上协同多种关系网络。因此,从垂直和平行两个角度将多种关系网络映射到同一个向量空间,提出一种基于多网络混合嵌入的服务推荐模型（MNHER）。首先,构建用户社交关系网络、服务标签共有网络、用户-服务异质信息网络;然后,通过多网络混合嵌入学习,得到用户和服务在同一向量空间的嵌入向量;最后,应用用户和服务的表征向量向目标用户推荐服务。此外,也对嵌入学习中的随机游走方法进行了优化,确保能更有效地提取和保留原网络的特征信息。为验证该方法的有效性,在三个公开数据集上与多种代表性的服务推荐方法进行了对比分析,相比基于单一关系网络和简单融合多关系网络的服务推荐方法,F-measure值分别可提高21%、15%。实验结果证明了多网络混合嵌入方法可有效地协同多关系网络,提高服务推荐质量。

关键词: 异质信息网络, 关系网络, 网络嵌入, 服务推荐, 协同过滤

Abstract:

The network embedding method can map the network nodes to a low-dimensional vector space and ext-ract the feature information of each node effectively. In the field of service recommendation, some studies show that the introduction of network embedding method can effectively alleviate the problem of data sparsity in the recom-mendation process. However, the existing network embedding methods are mostly aimed at a specific structure of the network, and do not cooperate with a variety of relationship networks from the source. Therefore, this paper proposes a service recommendation method based on multi-network hybrid embedding (MNHER), which maps mul-tiple relational networks to the same vector space from vertical and parallel perspectives. Firstly, the social network of users, the shared network of service tags and the user-service heterogeneous information network are constructed. Then, the hybrid embedding method proposed in this paper is used to obtain the embedding vector of users and services in the same vector space. Finally, the service recommendation is made to target users based on the embed-ding vector of users and services. In this paper, the random walk method is further optimized to extract and retain the characteristic information of the original network more effectively. In order to verify the effectiveness of the method proposed in this paper, it is compared with a variety of representative service recommendation methods on three public datasets, and the F-measure values of the service recommendation methods based on single relational network and simply fused multi-relational network are improved by 21% and 15%, respectively. It is proven that the method of multi-network hybrid embedding can effectively coordinate multi-relationship network and improve the quality of service recommendation.

Key words: heterogeneous information network, relational network, network embedding, service recommendation, collaborative filtering

中图分类号:

TP391

王雪纯, 吕晟凯, 吴浩, 何鹏, 曾诚. 多网络混合嵌入学习的服务推荐方法研究[J]. 计算机科学与探索, 2022, 16(7): 1529-1542.

WANG Xuechun, LYU Shengkai, WU Hao, HE Peng, ZENG Cheng. Research on Service Recommendation Method of Multi-network Hybrid Embed-ding Learning[J]. Journal of Frontiers of Computer Science and Technology, 2022, 16(7): 1529-1542.

图/表 11

图1 用户-服务关系网络

Fig.1 User-service network

图2 多网络混合嵌入学习模型

Fig.2 Multi-network hybrid embedding learning model

表1 实验数据集

Table 1 Experimental dataset

名称	用户	服务	评分数据	社交关系
Douban Movie	13 367	12 677	1 068 278	4 085
Yelp	16 239	14 284	198 397	158 590
Movielens	1 370	2 682	100 000	47 150

表2 信息网络数据统计情况

Table 2 Data statistics of information network

名称	网络	节点数	连边数
Douban Movie	用户同质信息网络	10 279	78 311
	服务同质信息网络	8 207	63 483
	用户社交网络	9 032	3 357
Yelp	同质信息网络	13 279	87 137
	服务同质信息网络	12 540	69 257
	用户社交网络	14 367	98 573
Movielens	用户同质信息网络	1 170	23 367
	服务同质信息网络	2 194	17 621
	用户社交网络	1 037	45 889

表3 不同模型在三个数据集上的推荐质量比较

Table 3 Comparison of recommendation quality of different models on three datasets

Dataset	Train rate	Metrics	UPCC	BPR	FM_HIN	Sorec	SemRec	MVIR	CPLR	MNHER
Yelp	80%	F1@10	0.148	0.124	0.189	0.180	0.216	0.295	0.317	0.396
		NDGG@10	0.146	0.132	0.190	0.190	0.206	0.244	0.270	0.271
		MAP@10	0.167	0.076	0.132	0.165	0.209	0.350	0.254	0.358
		MRR@10	0.393	0.190	0.273	0.260	0.382	0.503	0.442	0.552
	60%	F1@10	0.145	0.122	0.156	0.177	0.189	0.289	0.321	0.389
		NDGG@10	0.143	0.129	0.157	0.186	0.202	0.217	0.204	0.243
		MAP@10	0.151	0.075	0.067	0.064	0.117	0.164	0.221	0.289
		MRR@10	0.386	0.186	0.552	0.549	0.483	0.625	0.515	0.631
	40%	F1@10	0.136	0.118	0.149	0.171	0.184	0.234	0.311	0.377
		NDGG@10	0.138	0.125	0.182	0.181	0.195	0.240	0.216	0.258
		MAP@10	0.134	0.073	0.064	0.062	0.113	0.207	0.216	0.283
		MRR@10	0.374	0.181	0.526	0.532	0.468	0.530	0.534	0.680
	20%	F1@10	0.132	1.104	0.164	0.161	0.172	0.262	0.292	0.353
		NDGG@10	0.130	0.117	0.103	0.108	0.183	0.219	0.183	0.241
		MAP@10	0.083	0.068	0.060	0.058	0.106	0.113	0.220	0.263
		MRR@10	0.350	0.169	0.458	0.499	0.438	0.500	0.417	0.508
Douban Movie	80%	F1@10	0.108	0.119	0.152	0.150	0.170	0.260	0.345	0.319
		NDGG@10	0.118	0.043	0.257	0.272	0.173	0.222	0.393	0.428
		MAP@10	0.095	0.114	0.071	0.075	0.105	0.186	0.161	0.239
		MRR@10	0.494	0.242	0.439	0.447	0.552	0.605	0.524	0.655
	60%	F1@10	0.131	0.113	0.143	0.143	0.161	0.353	0.337	0.299
		NDGG@10	0.112	0.041	0.214	0.258	0.165	0.190	0.199	0.236
		MAP@10	0.072	0.108	0.051	0.072	0.100	0.170	0.135	0.194
		MRR@10	0.316	0.235	0.331	0.336	0.426	0.502	0.492	0.615
	40%	F1@10	0.086	0.106	0.156	0.149	0.160	0.293	0.296	0.338
		NDGG@10	0.100	0.192	0.052	0.054	0.163	0.199	0.345	0.406
		MAP@10	0.089	0.053	0.144	0.142	0.099	0.168	0.173	0.192
		MRR@10	0.384	0.225	0.365	0.416	0.513	0.632	0.554	0.693
	20%	F1@10	0.177	0.101	0.145	0.142	0.153	0.281	0.238	0.327
		NDGG@10	0.095	0.184	0.047	0.052	0.156	0.211	0.196	0.245
		MAP@10	0.085	0.051	0.130	0.136	0.095	0.169	0.147	0.184
		MRR@10	0.393	0.215	0.318	0.398	0.491	0.604	0.516	0.633
Movielens	80%	F1@10	0.105	0.112	0.111	0.119	0.143	0.209	0.205	0.282
		NDGG@10	0.109	0.105	0.103	0.108	0.102	0.179	0.206	0.257
		MAP@10	0.059	0.056	0.042	0.046	0.062	0.106	0.109	0.136
		MRR@10	0.169	0.181	0.208	0.293	0.238	0.293	0.272	0.469
	60%	F1@10	0.107	0.109	0.114	0.116	0.071	0.123	0.229	0.273
		NDGG@10	0.076	0.102	0.109	0.105	0.061	0.215	0.180	0.250
		MAP@10	0.033	0.054	0.047	0.045	0.030	0.068	0.125	0.132
		MRR@10	0.167	0.172	0.227	0.279	0.226	0.230	0.259	0.446
	40%	F1@10	0.136	0.104	0.106	0.111	0.068	0.194	0.220	0.262
		NDGG@10	0.109	0.098	0.094	0.103	0.059	0.138	0.192	0.239
		MAP@10	0.037	0.056	0.042	0.045	0.030	0.063	0.109	0.149
		MRR@10	0.159	0.207	0.216	0.227	0.224	0.291	0.237	0.442
	20%	F1@10	0.126	0.100	0.103	0.106	0.065	0.186	0.221	0.251
		NDGG@10	0.098	0.093	0.082	0.096	0.056	0.181	0.193	0.229
		MAP@10	0.046	0.047	0.050	0.038	0.025	0.091	0.091	0.113
		MRR@10	0.152	0.198	0.202	0.217	0.214	0.279	0.227	0.422

表3 不同模型在三个数据集上的推荐质量比较

Table 3 Comparison of recommendation quality of different models on three datasets

Dataset	Train rate	Metrics	UPCC	BPR	FM_HIN	Sorec	SemRec	MVIR	CPLR	MNHER
Yelp	80%	F1@10	0.148	0.124	0.189	0.180	0.216	0.295	0.317	0.396
		NDGG@10	0.146	0.132	0.190	0.190	0.206	0.244	0.270	0.271
		MAP@10	0.167	0.076	0.132	0.165	0.209	0.350	0.254	0.358
		MRR@10	0.393	0.190	0.273	0.260	0.382	0.503	0.442	0.552
	60%	F1@10	0.145	0.122	0.156	0.177	0.189	0.289	0.321	0.389
		NDGG@10	0.143	0.129	0.157	0.186	0.202	0.217	0.204	0.243
		MAP@10	0.151	0.075	0.067	0.064	0.117	0.164	0.221	0.289
		MRR@10	0.386	0.186	0.552	0.549	0.483	0.625	0.515	0.631
	40%	F1@10	0.136	0.118	0.149	0.171	0.184	0.234	0.311	0.377
		NDGG@10	0.138	0.125	0.182	0.181	0.195	0.240	0.216	0.258
		MAP@10	0.134	0.073	0.064	0.062	0.113	0.207	0.216	0.283
		MRR@10	0.374	0.181	0.526	0.532	0.468	0.530	0.534	0.680
	20%	F1@10	0.132	1.104	0.164	0.161	0.172	0.262	0.292	0.353
		NDGG@10	0.130	0.117	0.103	0.108	0.183	0.219	0.183	0.241
		MAP@10	0.083	0.068	0.060	0.058	0.106	0.113	0.220	0.263
		MRR@10	0.350	0.169	0.458	0.499	0.438	0.500	0.417	0.508
Douban Movie	80%	F1@10	0.108	0.119	0.152	0.150	0.170	0.260	0.345	0.319
		NDGG@10	0.118	0.043	0.257	0.272	0.173	0.222	0.393	0.428
		MAP@10	0.095	0.114	0.071	0.075	0.105	0.186	0.161	0.239
		MRR@10	0.494	0.242	0.439	0.447	0.552	0.605	0.524	0.655
	60%	F1@10	0.131	0.113	0.143	0.143	0.161	0.353	0.337	0.299
		NDGG@10	0.112	0.041	0.214	0.258	0.165	0.190	0.199	0.236
		MAP@10	0.072	0.108	0.051	0.072	0.100	0.170	0.135	0.194
		MRR@10	0.316	0.235	0.331	0.336	0.426	0.502	0.492	0.615
	40%	F1@10	0.086	0.106	0.156	0.149	0.160	0.293	0.296	0.338
		NDGG@10	0.100	0.192	0.052	0.054	0.163	0.199	0.345	0.406
		MAP@10	0.089	0.053	0.144	0.142	0.099	0.168	0.173	0.192
		MRR@10	0.384	0.225	0.365	0.416	0.513	0.632	0.554	0.693
	20%	F1@10	0.177	0.101	0.145	0.142	0.153	0.281	0.238	0.327
		NDGG@10	0.095	0.184	0.047	0.052	0.156	0.211	0.196	0.245
		MAP@10	0.085	0.051	0.130	0.136	0.095	0.169	0.147	0.184
		MRR@10	0.393	0.215	0.318	0.398	0.491	0.604	0.516	0.633
Movielens	80%	F1@10	0.105	0.112	0.111	0.119	0.143	0.209	0.205	0.282
		NDGG@10	0.109	0.105	0.103	0.108	0.102	0.179	0.206	0.257
		MAP@10	0.059	0.056	0.042	0.046	0.062	0.106	0.109	0.136
		MRR@10	0.169	0.181	0.208	0.293	0.238	0.293	0.272	0.469
	60%	F1@10	0.107	0.109	0.114	0.116	0.071	0.123	0.229	0.273
		NDGG@10	0.076	0.102	0.109	0.105	0.061	0.215	0.180	0.250
		MAP@10	0.033	0.054	0.047	0.045	0.030	0.068	0.125	0.132
		MRR@10	0.167	0.172	0.227	0.279	0.226	0.230	0.259	0.446
	40%	F1@10	0.136	0.104	0.106	0.111	0.068	0.194	0.220	0.262
		NDGG@10	0.109	0.098	0.094	0.103	0.059	0.138	0.192	0.239
		MAP@10	0.037	0.056	0.042	0.045	0.030	0.063	0.109	0.149
		MRR@10	0.159	0.207	0.216	0.227	0.224	0.291	0.237	0.442
	20%	F1@10	0.126	0.100	0.103	0.106	0.065	0.186	0.221	0.251
		NDGG@10	0.098	0.093	0.082	0.096	0.056	0.181	0.193	0.229
		MAP@10	0.046	0.047	0.050	0.038	0.025	0.091	0.091	0.113
		MRR@10	0.152	0.198	0.202	0.217	0.214	0.279	0.227	0.422

图3 在不同数据集上与基线的性能对比

Fig.3 Performance comparisons with baselines on different datasets

表4 考虑不同关系的实验结果对比

Table 4 Comparison of experimental results considering different relations

Method	Yelp				Douban Movie				Movielens
Method	F1	NDGG	MAP	MRR	F1	NDGG	MAP	MRR	F1	NDGG	MAP	MRR
$M N H E R n u l l$	0.339	0.255	0.306	0.470	0.310	0.381	0.206	0.653	0.239	0.220	0.118	0.346
$M N H E R s o c i a l$	0.354	0.267	0.320	0.491	0.329	0.368	0.216	0.683	0.250	0.230	0.123	0.392
$M N H E R c o t a g$	0.382	0.287	0.370	0.546	0.326	0.412	0.203	0.659	0.272	0.258	0.140	0.427
$M N H E R$	0.396	0.285	0.376	0.580	0.345	0.380	0.251	0.677	0.096	0.270	0.133	0.493

表4 考虑不同关系的实验结果对比

Table 4 Comparison of experimental results considering different relations

Method	Yelp				Douban Movie				Movielens
Method	F1	NDGG	MAP	MRR	F1	NDGG	MAP	MRR	F1	NDGG	MAP	MRR
$M N H E R n u l l$	0.339	0.255	0.306	0.470	0.310	0.381	0.206	0.653	0.239	0.220	0.118	0.346
$M N H E R s o c i a l$	0.354	0.267	0.320	0.491	0.329	0.368	0.216	0.683	0.250	0.230	0.123	0.392
$M N H E R c o t a g$	0.382	0.287	0.370	0.546	0.326	0.412	0.203	0.659	0.272	0.258	0.140	0.427
$M N H E R$	0.396	0.285	0.376	0.580	0.345	0.380	0.251	0.677	0.096	0.270	0.133	0.493

图4 考虑不同关系的实验结果对比

Fig.4 Comparison of experimental results considering different relations

图5 参数 δ的实验分析结果

Fig.5 Experimental analysis of parameter δ

表5 不同中心性度量方法下的实验结果

Table 5 Experimental results under different centrality measurement methods

Method	Yelp				Douban Movie				Movielens
Method	F1	NDGG	MAP	MRR	F1	NDGG	MAP	MRR	F1	NDGG	MAP	MRR
NULL	0.322	0.243	0.291	0.447	0.295	0.267	0.196	0.621	0.227	0.209	0.141	0.329
BC	0.339	0.255	0.306	0.470	0.310	0.281	0.206	0.653	0.239	0.220	0.148	0.346
EC	0.275	0.261	0.336	0.496	0.297	0.375	0.185	0.599	0.247	0.298	0.128	0.400
PR	0.416	0.285	0.376	0.480	0.330	0.450	0.251	0.688	0.296	0.270	0.143	0.493

图6 参数mw、p实验分析

Fig.6 Experimental analysis of parameter mw, p

参考文献 40

[1]	BURKE R. Hybrid recommender systems: survey and experi-ments[J]. User Modeling and User-Adapted Interaction, 2002, 12(4): 331-370. DOI URL
[2]	ZHOU P Y, LIU J, LIU X, et al. Is deep learning better than traditional approaches in tag recommendation for software information sites?[J]. Information and Software Technology, 2019, 109: 1-13. DOI URL
[3]	CAI X Y, HAN J W, PAN S R, et al. Heterogeneous infor-mation network embedding based personalized query-focused astronomy reference paper recommendation[J]. International Journal of Computational Intelligence Systems, 2018, 11(1): 591-599. DOI URL
[4]	WEI S X, ZHENG X L, CHEN D R, et al. A hybrid approach for movie recommendation via tags and ratings[J]. Electronic Commerce Research and Applications, 2016, 18: 83-94. DOI URL
[5]	LECUN Y, BENGIO Y, HINTON G, et al. Deep learning[J]. Nature, 2015, 521(7553): 436-444. DOI URL
[6]	ZHANG R, XIE P, WANG C, et al. Classifying transporta-tion mode and speed from trajectory data via deep multi-scale learning[J]. Computer Networks, 2019, 162: 106861. DOI URL
[7]	JIAO P F, TANG M H, LIU H T, et al. Variational auto enco-der based bipartite network embedding by integrating local and global structure[J]. Information Sciences, 2020, 519: 9-21. DOI URL
[8]	GROVER A, LESKOVEC J. node2vec: scalable feature lear-ning for networks[C]// Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Francisco, Aug 13-17, 2016. New York: ACM, 2016: 855-864.
[9]	TANG J, QU M, WANG M Z, et al. LINE: large-scale infor-mation network embedding[C]// Proceedings of the 24th Inter-national Conference on World Wide Web, Florence, May 18-22, 2015. New York: ACM, 2015: 1067-1077.
[10]	WANG D X, CUI P, ZHU W W. Structural deep network embedding[C]// Proceedings of the 22nd ACM SIGKDD Inter-national Conference on Knowledge Discovery and Data Mining, San Francisco, Aug 13-17, 2016. New York: ACM, 2016: 1225-1234.
[11]	GAO M, CHEN L H, HE X N, et al. BiNE: bipartite network embedding[C]// Proceedings of the 41st International ACM SIGIR Conference on Research & Development in Informa-tion Retrieval, Ann Arbor, Jul 8-12, 2018. New York: ACM, 2018: 715-724.
[12]	WU H, PENG Y N, HE P, et al. Service recommendation based on the use of reliable user network supplementary tags[J]. Internet of Things Technology, 2019, 9(2): 48-51.
[13]	WANG Z K. Research on recommendation algorithm based on representation learning in heterogeneous information net-works[D]. Beijing: Peking University, 2019.
[14]	LUO L, XIE H R, RAO Y H, et al. Personalized recommen-dation by matrix co-factorization with tags and time informa-tion[J]. Expert Systems with Application, 2019, 119: 311-321. DOI URL
[15]	SHI W S, LIU X M, YU Q. Correlation-aware multi-label active learning for Web service tag recommendation[C]// Proceedings of the 2017 IEEE International Conference on Web Services, Honolulu, Jun 25-30, 2017. Piscataway: IEEE, 2017: 229-236.
[16]	ZHANG S, XIA Y, LI X. Diversified recommendation algo-rithm for hybrid label based on matrix factorization[C]// Proceedings of the 2018 IEEE 3rd International Conference on Cloud Computing and Big Data Analysis, Chengdu, Apr 20-22, 2018. Piscataway: IEEE, 2018: 39-44.
[17]	QIANG H, YAN G. A method of personalized recommen-dation based on multi-label propagation for overlapping com-munity detection[C]// Proceedings of the 2012 3rd Interna-tional Conference on System Science, Engineering Design and Manufacturing Informatization, Chengdu, Oct 20-21, 2012. Piscataway: IEEE, 2012: 360-364.
[18]	BA Q L, LI X Y, BAI Z Y. A similarity calculating approach simulated from TF-IDF in collaborative filtering recommen-dation[C]// Proceedings of the 5th International Conference on Multimedia Information Networking & Security, Nov 1-3, 2013. New York: ACM, 2013: 738-741.
[19]	OTSUKA E, WALLACE S A, CHIU D. Design and evalua-tion of a Twitter hashtag recommendation system[C]// Procee-dings of the 18th International Database Engineering & Appli-cations Symposium, Porto, Jul 7-9, 2014. New York: ACM, 2014: 330-333.
[20]	WANG X, LU W, ESTER M. Social recommendation with strong and weak ties[C]// Proceedings of the 25th ACM Inter-national Conference on Information and Knowledge Mana-gement, Indianapolis, Oct 24-28, 2016. New York: ACM, 2016: 5-14.
[21]	ZHOU P, ZHOU Y X, Wu D P, et al. Differentially private online learning for cloud-based video recommendation with multimedia big data in social networks[J]. IEEE Transac-tions on Multimedia, 2016, 18(6): 1217-1229.
[22]	HU Y, YI X, DAVIS L S. Collaborative fashion recommen-dation: a functional tensor factorization approach[C]// Procee-dings of the 23rd Annual ACM Conference on Multimedia Conference, Brisbane, Oct 26-30, 2015. New York: ACM, 2015: 129-138.
[23]	YANG B, LEI Y, LIU J, et al. Social collaborative filtering by trust[J]. IEEE Transactions on Pattern Analysis and Mac-hine Intelligence, 2017, 39(8): 1633-1647.
[24]	JIANG Z G, ZHOU A, WANG S G, et al. Personalized service recommendation for collaborative tagging systems with social relations and temporal influences[C]// Procee-dings of the 2016 IEEE International Conference on Services Computing, San Francisco, Jun 27-Jul 2, 2016. Washington: IEEE Computer Society, 2016: 786-789.
[25]	LIU Y, TIAN Z, SUN J, et al. Distributed representation learning via node2vec for implicit feedback recommenda-tion[J]. Neural Computing and Applications, 2019, 7: 1-11. DOI URL
[26]	ZHUANG F, ZHANG Z, QIAN M, et al. Representation learning via dual-autoencoder for recommendation[J]. Neural Networks, 2017, 90: 83-89. DOI URL
[27]	吴玺煜, 陈启买, 刘海, 等. 基于知识图谱表示学习的协同过滤推荐算法[J]. 计算机工程, 2018, 44(2): 226-232.
	WU X Y, CHEN Q M, LIU H, et al. Collaborative filtering recommendation algorithm based on representation learning of knowledge graph[J]. Computer Engineering, 2018, 44(2): 226-232.
[28]	WU L, QUAN C, LI C, et al. A context-aware user-item representation learning for item recommendation[J]. Transac-tions on Information Systems, 2019, 37(2): 1-29.
[29]	KONG X, MAO M, WANG W, et al. VOPRec: vector repre-sentation learning of papers with text information and struc-tural identity for recommendation[J]. IEEE Transactions on Emerging Topics in Computing, 2021, 9(1): 226-237. DOI URL
[30]	SALAH A, LAUW H W. Probabilistic collaborative repre-sentation learning for personalized item recommendation[C]// Proceedings of the 34th Conference on Uncertainty in A.pngicial Intelligence, Monterey, Aug 6-10, 2018: 998-1008.
[31]	HAN X T, SHI C, ZHENG L, et al. Representation learning with depth and breadth for recommendation using multi-view data[C]// LNCS 10987: Proceedings of the 2nd Interna-tional Joint Conference on Web and Big Data, Macau, China, Jul 23-25, 2018. Cham: Springer, 2018: 181-188.
[32]	SHI C, HU B, ZHAO X, et al. Heterogeneous information network embedding for recommendation[J]. IEEE Transac-tions on Knowledge & Data Engineering, 2017, 31(2): 357-370.
[33]	MIKOLOV T, CHEN K, CORRADO G, et al. Efficient estimation of word representations in vector space[J]. arXiv:1301. 3781, 2013.
[34]	PAGE L, BRIN S, MOTWANI R, et al. The PageRank cita-tion ranking: bringing order to the web[R]. Stanford: Stan-ford InfoLab., 1999: 161-172.
[35]	MA H, YANG H X, LYU M R, et al. SoRec: social recom-mendation using probabilistic matrix factorization[C]// Procee-dings of the 17th ACM Conference on Information and Know-ledge Management, Napa Valley, Oct 26-30, 2008. New York: ACM, 2008: 931-940.
[36]	SHI C, ZHANG Z Q, LUO P, et al. Semantic path based personalized recommendation on weighted heterogeneous information networks[C]// Proceedings of the 24th ACM Inter-national Conference on Information and Knowledge Mana-gement, Melbourne, Oct 19-23, 2015. New York: ACM, 2015: 453-462.
[37]	RENDLE S, FREUDENTHALER C, GANTNER Z, et al. BPR: Bayesian personalized ranking from implicit feedback[C]// Proceedings of the 25th Conference on Uncertainty in A.pngicial Intelligence, Montreal, Jun 18-21, 2009: 452-461.
[38]	LIU H Z, WU Z H, ZHANG X. CPLR: collaborative pair-wise learning to rank for personalized recommendation[J]. Knowledge-Based Systems, 2018, 148: 31-40. DOI URL
[39]	SHAO L S, ZHANG J, WEI Y, et al. Personalized QoS predi-ction for Web services via collaborative filtering[C]// Procee-dings of the 2007 IEEE International Conference on Web Services, Salt Lake City, Jul 9-13, 2007. Washington: IEEE Computer Society, 2007: 439-446.
[40]	RENDLE S. Factorization machines with libFM[J]. ACM Transactions on Intelligent Systems and Technology, 2012, 3(3): 57.

多网络混合嵌入学习的服务推荐方法研究

Research on Service Recommendation Method of Multi-network Hybrid Embed-ding Learning

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 11

参考文献 40

相关文章 15

编辑推荐

Metrics

[1]	武森, 董雅贤, 魏桂英, 高晓楠. 面向稀疏数据的协同过滤用户相似度计算研究[J]. 计算机科学与探索, 2022, 16(5): 1043-1052.
[2]	张全贵, 胡嘉燕, 王丽. 耦合用户公共特征的单类协同过滤推荐算法[J]. 计算机科学与探索, 2022, 16(3): 637-648.
[3]	吴正洋, 汤庸, 刘海. 个性化学习推荐研究综述[J]. 计算机科学与探索, 2022, 16(1): 21-40.
[4]	陈洁, 陈嘉琳, 赵姝, 张燕平. 层次标签引导的属性网络嵌入[J]. 计算机科学与探索, 2021, 15(7): 1279-1288.
[5]	赵传，张凯涵，梁吉业. 非对称的异质信息网络推荐算法[J]. 计算机科学与探索, 2020, 14(6): 939-946.
[6]	邢长征，赵宏宝，张全贵，郭亚兰. 融合评论文本层级注意力和外积的推荐方法[J]. 计算机科学与探索, 2020, 14(6): 947-957.
[7]	姜芸，何伟，崔立真，杨倩，刘磊. 移动情景和用户轨迹感知的众包服务推荐[J]. 计算机科学与探索, 2019, 13(9): 1471-1480.
[8]	周慧，赵中英，李超. 面向异质信息网络的表示学习方法研究综述[J]. 计算机科学与探索, 2019, 13(7): 1081-1093.
[9]	王宇琛，王宝亮，侯永宏. 融合协同过滤与上下文信息的Bandits推荐算法[J]. 计算机科学与探索, 2019, 13(3): 361-373.
[10]	邵礼旭，段玉聪，周长兵，高洪皓，陈世展. 数据、信息和知识三层图谱架构的推荐服务设计[J]. 计算机科学与探索, 2019, 13(2): 214-225.
[11]	朱宏伟，游晓明，刘升. 协同过滤策略的异构双种群蚁群算法[J]. 计算机科学与探索, 2019, 13(10): 1754-1767.
[12]	张以文，艾晓飞，崔光明，钱付兰. 联合用户兴趣矩阵及全局偏好的推荐算法[J]. 计算机科学与探索, 2018, 12(2): 197-207.
[13]	郭宁宁，王宝亮，侯永宏，常鹏. 融合社交网络特征的协同过滤推荐算法[J]. 计算机科学与探索, 2018, 12(2): 208-217.
[14]	冯勇，徐红艳，王嵘冰，郭浩. 融入项目相关性的加权Slope One算法研究[J]. 计算机科学与探索, 2018, 12(10): 1691-1700.
[15]	房倩琦，柳玲，文俊浩，曾骏，高旻. 社交关系在基于模型社会化推荐系统中的影响[J]. 计算机科学与探索, 2018, 12(1): 82-91.