《计算机应用》唯一官方网站 ›› 2025, Vol. 45 ›› Issue (8): 2592-2599.DOI: 10.11772/j.issn.1001-9081.2024071038
• 数据科学与技术 • 上一篇
收稿日期:
2024-07-23
修回日期:
2024-10-11
接受日期:
2024-10-11
发布日期:
2024-11-19
出版日期:
2025-08-10
通讯作者:
马应龙
作者简介:
王义(2000—),男,河北沧州人,硕士研究生,主要研究方向:社交推荐、图嵌入
基金资助:
Received:
2024-07-23
Revised:
2024-10-11
Accepted:
2024-10-11
Online:
2024-11-19
Published:
2025-08-10
Contact:
Yinglong MA
About author:
WANG Yi, born in 2000, M. S. candidate. His research interests include social recommendation, graph embedding.
Supported by:
摘要:
除了考虑用户间的社交关系,挖掘项间隐含的关系特征对提升用户和项的表示学习能力同样具有至关重要的作用。当前社交项推荐中静态项图构建过程难以准确抓取项间的潜在关系,且后续的图融合过程缺乏深度交互,这限制了相关模型对多图特征间复杂且多层次关系的理解能力。因此,提出一种基于项图动态适应性生成的多任务社交项推荐方法(MTDAG)。首先,在基于多任务学习(MTL)的联合训练中,使用项图动态生成模块结合下游推荐任务的反馈信息适应性地调整项图结构;其次,使用社交项推荐模块通过深层次的多图特征融合方法在各输入图之间迭代地传播用户和项的特征表示;最后,在Yelp和Ciao两个公共数据集上把MTDAG与ECGN (Efficient Complementary Graph convolutional Network)和MGL (Meta Graph Learning framework)等6种基线方法比较。实验结果表明,MTDAG在命中率(HR)、召回率(Recall)和归一化折损累计增益(NDCG)上均至少提高了3%且MTDAG的鲁棒性在针对冷用户和冷项推荐的评估实验中得到了充分验证,实验结果表明,MTDAG可以在一定程度上解决交互稀疏的冷用户和冷项的推荐问题。
中图分类号:
王义, 马应龙. 基于项图动态适应性生成的多任务社交项推荐方法[J]. 计算机应用, 2025, 45(8): 2592-2599.
Yi WANG, Yinglong MA. Multi-task social item recommendation method based on dynamic adaptive generation of item graph[J]. Journal of Computer Applications, 2025, 45(8): 2592-2599.
数据集 | 用户数 | 项数 | 交互边数 | 社交数 | 交互密度/% |
---|---|---|---|---|---|
Yelp | 17 237 | 38 342 | 204 448 | 143 765 | 0.030 9 |
Ciao | 7 373 | 91 091 | 226 307 | 111 746 | 0.033 7 |
表1 数据集统计信息
Tab. 1 Statistics of datasets
数据集 | 用户数 | 项数 | 交互边数 | 社交数 | 交互密度/% |
---|---|---|---|---|---|
Yelp | 17 237 | 38 342 | 204 448 | 143 765 | 0.030 9 |
Ciao | 7 373 | 91 091 | 226 307 | 111 746 | 0.033 7 |
数据集 | 方法 | HR@10 | HR@20 | HR@30 | Recall@10 | Recall@20 | Recall@30 | NDCG@10 | NDCG@20 | NDCG@30 |
---|---|---|---|---|---|---|---|---|---|---|
Yelp | NGCF | 0.035 17 | 0.061 64 | 0.084 14 | 0.024 33 | 0.042 93 | 0.059 31 | 0.013 57 | 0.018 66 | 0.022 44 |
LightGCN | 0.038 58 | 0.065 36 | 0.087 06 | 0.026 72 | 0.045 67 | 0.061 00 | 0.014 45 | 0.019 31 | 0.023 09 | |
Diffnet | 0.033 99 | 0.059 69 | 0.075 07 | 0.022 47 | 0.039 31 | 0.057 62 | 0.012 10 | 0.017 17 | 0.020 05 | |
SocialLGN | 0.038 95 | 0.066 25 | 0.026 93 | 0.045 60 | 0.014 72 | 0.019 74 | ||||
ECGN | 0.088 23 | 0.060 74 | 0.023 14 | |||||||
MGL | 0.034 44 | 0.060 90 | 0.083 72 | 0.024 54 | 0.043 72 | 0.060 31 | 0.013 51 | 0.018 75 | 0.022 58 | |
MTDAG | 0.040 83 | 0.069 42 | 0.091 36 | 0.028 35 | 0.048 57 | 0.064 30 | 0.015 43 | 0.021 04 | 0.024 51 | |
Ciao | NGCF | 0.090 42 | 0.130 15 | 0.160 14 | 0.038 67 | 0.056 26 | 0.070 88 | 0.026 30 | 0.031 64 | 0.035 86 |
LightGCN | 0.091 76 | 0.130 95 | 0.161 74 | 0.039 54 | 0.057 32 | 0.073 12 | 0.026 53 | 0.032 21 | 0.036 51 | |
Diffnet | 0.090 20 | 0.127 99 | 0.150 23 | 0.037 24 | 0.056 64 | 0.069 16 | 0.027 04 | 0.032 87 | 0.036 23 | |
SocialLGN | 0.095 94 | 0.139 06 | 0.040 49 | 0.062 11 | 0.078 54 | 0.028 48 | 0.034 26 | 0.037 90 | ||
ECGN | 0.170 23 | |||||||||
MGL | 0.089 61 | 0.124 04 | 0.153 44 | 0.036 59 | 0.054 28 | 0.067 19 | 0.025 62 | 0.029 13 | 0.033 20 | |
MTDAG | 0.100 11 | 0.147 10 | 0.177 96 | 0.042 61 | 0.064 32 | 0.081 58 | 0.029 54 | 0.036 52 | 0.040 93 |
表2 MTDAG和基线方法在两个数据集上的实验结果
Tab. 2 Experimental results of MTDAG and baseline methods on two datasets
数据集 | 方法 | HR@10 | HR@20 | HR@30 | Recall@10 | Recall@20 | Recall@30 | NDCG@10 | NDCG@20 | NDCG@30 |
---|---|---|---|---|---|---|---|---|---|---|
Yelp | NGCF | 0.035 17 | 0.061 64 | 0.084 14 | 0.024 33 | 0.042 93 | 0.059 31 | 0.013 57 | 0.018 66 | 0.022 44 |
LightGCN | 0.038 58 | 0.065 36 | 0.087 06 | 0.026 72 | 0.045 67 | 0.061 00 | 0.014 45 | 0.019 31 | 0.023 09 | |
Diffnet | 0.033 99 | 0.059 69 | 0.075 07 | 0.022 47 | 0.039 31 | 0.057 62 | 0.012 10 | 0.017 17 | 0.020 05 | |
SocialLGN | 0.038 95 | 0.066 25 | 0.026 93 | 0.045 60 | 0.014 72 | 0.019 74 | ||||
ECGN | 0.088 23 | 0.060 74 | 0.023 14 | |||||||
MGL | 0.034 44 | 0.060 90 | 0.083 72 | 0.024 54 | 0.043 72 | 0.060 31 | 0.013 51 | 0.018 75 | 0.022 58 | |
MTDAG | 0.040 83 | 0.069 42 | 0.091 36 | 0.028 35 | 0.048 57 | 0.064 30 | 0.015 43 | 0.021 04 | 0.024 51 | |
Ciao | NGCF | 0.090 42 | 0.130 15 | 0.160 14 | 0.038 67 | 0.056 26 | 0.070 88 | 0.026 30 | 0.031 64 | 0.035 86 |
LightGCN | 0.091 76 | 0.130 95 | 0.161 74 | 0.039 54 | 0.057 32 | 0.073 12 | 0.026 53 | 0.032 21 | 0.036 51 | |
Diffnet | 0.090 20 | 0.127 99 | 0.150 23 | 0.037 24 | 0.056 64 | 0.069 16 | 0.027 04 | 0.032 87 | 0.036 23 | |
SocialLGN | 0.095 94 | 0.139 06 | 0.040 49 | 0.062 11 | 0.078 54 | 0.028 48 | 0.034 26 | 0.037 90 | ||
ECGN | 0.170 23 | |||||||||
MGL | 0.089 61 | 0.124 04 | 0.153 44 | 0.036 59 | 0.054 28 | 0.067 19 | 0.025 62 | 0.029 13 | 0.033 20 | |
MTDAG | 0.100 11 | 0.147 10 | 0.177 96 | 0.042 61 | 0.064 32 | 0.081 58 | 0.029 54 | 0.036 52 | 0.040 93 |
[1] | FAN W, MA Y, LI Q, et al. Graph neural networks for social recommendation[C]// Proceedings of the 2019 World Wide Web Conference. New York: ACM, 2019: 417-426. |
[2] | McPHERSON M, SMITH-LOVIN L, COOK J M. Birds of a feather: homophily in social networks[J]. Annual Review of Sociology, 2001, 27: 415-444. |
[3] | ANAGNOSTOPOULOS A, KUMAR R, MAHDIAN M. Influence and correlation in social networks[C]// Proceedings of the 14th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York: ACM, 2008: 7-15. |
[4] | WANG S, HU L, WANG Y, et al. Graph learning based recommender systems: a review[C]// Proceedings of the 30th International Joint Conference on Artificial Intelligence. California: ijcai.org, 2021: 4644-4652. |
[5] | CHEN J, XIN X, LIANG X, et al. GDSRec: graph-based decentralized collaborative filtering for social recommendation[J]. IEEE Transactions on Knowledge and Data Engineering, 2023, 35(5): 4813-4824. |
[6] | YANG B, LEI Y, LIU J, et al. Social collaborative filtering by trust[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(8): 1633-1647. |
[7] | YU X, TU L, CHAI L, et al. Construction of implicit social network and recommendation between users and items via the ISR-RRM algorithm[J]. Expert Systems with Applications, 2024, 235: No.121229. |
[8] | HAN J, TAO Q, TANG Y, et al. DH-HGCN: dual homogeneity hypergraph convolutional network for multiple social recommendations[C]// Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM, 2022: 2190-2194. |
[9] | WU B, ZHONG L, LI H, et al. Efficient complementary graph convolutional network without negative sampling for item recommendation[J]. Knowledge-Based Systems, 2022, 256: No.109758. |
[10] | FAN W, MA Y, LI Q, et al. A graph neural network framework for social recommendations[J]. IEEE Transactions on Knowledge and Data Engineering, 2022, 34(5): 2033-2047. |
[11] | WEI C, LIANG J, LIU D, et al. Meta graph learning for long-tail recommendation[C]// Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining. New York: ACM, 2023: 2512-2522. |
[12] | LIAO J, ZHOU W, LUO F, et al. SocialLGN: light graph convolution network for social recommendation[J]. Information Sciences, 2022, 589: 595-607. |
[13] | WU L, SUN P, FU Y, et al. A neural influence diffusion model for social recommendation[C]// Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM, 2019: 235-244. |
[14] | LIU Y, CHEN L, HE X, et al. Modelling high-order social relations for item recommendation[J]. IEEE Transactions on Knowledge and Data Engineering, 2022, 34(9): 4385-4397. |
[15] | YANG L, LIU Z, DOU Y, et al. ConsisRec: enhancing GNN for social recommendation via consistent neighbor aggregation[C]// Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM, 2021: 2141-2145. |
[16] | WANG L, ZHOU W, LIU L, et al. Deep adaptive collaborative graph neural network for social recommendation[J]. Expert Systems with Applications, 2023, 229(Pt A): No.120410. |
[17] | JIANG W, GAO X, XU G, et al. Challenging low homophily in social recommendation[C]// Proceedings of the ACM Web Conference 2024. New York: ACM, 2024: 3476-3484. |
[18] | DU J, YE Z, YAO L, et al. Socially-aware dual contrastive learning for cold-start recommendation[C]// Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM, 2022: 1927-1932. |
[19] | FILHO R J R, WEHRMANN J, BARROS R C. Leveraging deep visual features for content-based movie recommender systems[C]// Proceedings of the 2017 International Joint Conference on Neural Networks. Piscataway: IEEE, 2017: 604-611. |
[20] | ZHOU Z, ZHANG L, YANG N. Contrastive collaborative filtering for cold-start item recommendation[C]// Proceedings of the ACM Web Conference 2023. New York: ACM, 2023: 928-937. |
[21] | HUANG F, WANG Z, HUANG X, et al. Aligning distillation for cold-start item recommendation[C]// Proceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM, 2023: 1147-1157. |
[22] | ZHU Y, XIE R, ZHUANG F, et al. Learning to warm up cold item embeddings for cold-start recommendation with meta scaling and shifting networks[C]// Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM, 2021: 1167-1176. |
[23] | ZHOU W, HAQ A U, QIU L, et al. Multi-view social recommendation via matrix factorization with sub-linear convergence rate[J]. Expert Systems with Applications, 2024, 237(Pt C): No.121687. |
[24] | ZHANG M, YIN R, YANG Z, et al. Advances and challenges of multi-task learning method in recommender system: a survey[EB/OL]. [2024-03-17].. |
[25] | FENG X, LIU Z, WU W, et al. Social recommendation via deep neural network-based multi-task learning[J]. Expert Systems with Applications, 2022, 206: No.117755. |
[26] | ZHANG Y, CHENG D Z, YAO T, et al. A model of two tales: dual transfer learning framework for improved long-tail item recommendation[C]// Proceedings of the Web Conference 2021. New York: ACM, 2021: 2220-2231. |
[27] | CHEN H, WANG Z, HUANG F, et al. Generative adversarial framework for cold-start item recommendation[C]// Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM, 2022: 2565-2571. |
[28] | YU D, ZHANG R, JIANG Z, et al. Graph-revised convolutional network[C]// Proceedings of the 2020 European Conference on Machine Learning and Knowledge Discovery in Databases, LNCS 12459. Cham: Springer, 2021: 378-393. |
[29] | HE X, DENG K, WANG X, et al. LightGCN: simplifying and powering graph convolution network for recommendation[C]// Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM, 2020: 639-648. |
[30] | WANG X, HE X, WANG M, et al. Neural graph collaborative filtering[C]// Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM, 2019: 165-174. |
[31] | KINGMA D P, BA J L. Adam: a method for stochastic optimization[EB/OL]. [2024-03-23].. |
[32] | LI J, LU K, HUANG Z, et al. On both cold-start and long-tail recommendation with social data[J]. IEEE Transactions on Knowledge and Data Engineering, 2021, 33(1): 194-208. |
[1] | 涂银川, 郭勇, 毛恒, 任怡, 张建锋, 李宝. 基于分布式环境的图神经网络模型训练效率与训练性能评估[J]. 《计算机应用》唯一官方网站, 2025, 45(8): 2409-2420. |
[2] | 赵彪, 秦玉华, 田荣坤, 胡月航, 陈芳锐. 依赖类型及距离增强的方面级情感分析模型[J]. 《计算机应用》唯一官方网站, 2025, 45(8): 2507-2514. |
[3] | 梁辰, 王奕森, 魏强, 杜江. 基于Tsransformer-GCN的源代码漏洞检测方法[J]. 《计算机应用》唯一官方网站, 2025, 45(7): 2296-2303. |
[4] | 张子墨, 赵雪专. 多尺度稀疏图引导的视觉图神经网络[J]. 《计算机应用》唯一官方网站, 2025, 45(7): 2188-2194. |
[5] | 陈丹阳, 张长伦. 多尺度去相关的图卷积网络模型[J]. 《计算机应用》唯一官方网站, 2025, 45(7): 2180-2187. |
[6] | 张悦岚, 苏静, 赵航宇, 杨白利. 基于知识感知与交互的多视图蒸馏推荐算法[J]. 《计算机应用》唯一官方网站, 2025, 45(7): 2211-2220. |
[7] | 姜超英, 李倩, 刘宁, 刘磊, 崔立真. 基于图对比学习的再入院预测模型[J]. 《计算机应用》唯一官方网站, 2025, 45(6): 1784-1792. |
[8] | 闫龙博, 毛文涛, 仲志鸿, 范黎林. 面向城市排水管网缺陷诊断的鲁棒无监督多任务异常检测方法[J]. 《计算机应用》唯一官方网站, 2025, 45(6): 1833-1840. |
[9] | 游兰, 张雨昂, 刘源, 陈智军, 王伟, 曾星, 何张玮. 基于协作贡献网络的开源项目开发者推荐[J]. 《计算机应用》唯一官方网站, 2025, 45(4): 1213-1222. |
[10] | 党伟超, 宋楚君, 高改梅, 刘春霞. 基于级联残差图卷积网络的多行为推荐[J]. 《计算机应用》唯一官方网站, 2025, 45(4): 1223-1231. |
[11] | 党伟超, 温鑫瑜, 高改梅, 刘春霞. 基于多视图多尺度对比学习的图协同过滤[J]. 《计算机应用》唯一官方网站, 2025, 45(4): 1061-1068. |
[12] | 王聪, 史艳翠. 基于多视角学习的图神经网络群组推荐模型[J]. 《计算机应用》唯一官方网站, 2025, 45(4): 1205-1212. |
[13] | 田仁杰, 景明利, 焦龙, 王飞. 基于混合负采样的图对比学习推荐算法[J]. 《计算机应用》唯一官方网站, 2025, 45(4): 1053-1060. |
[14] | 袁成哲, 陈国华, 李丁丁, 朱源, 林荣华, 钟昊, 汤庸. ScholatGPT:面向学术社交网络的大语言模型及智能应用[J]. 《计算机应用》唯一官方网站, 2025, 45(3): 755-764. |
[15] | 马汉达, 吴亚东. 多域时空层次图神经网络的空气质量预测[J]. 《计算机应用》唯一官方网站, 2025, 45(2): 444-452. |
阅读次数 | ||||||
全文 |
|
|||||
摘要 |
|
|||||