赞
踩
本文基于两篇综述来介绍基于图神经网络的推荐系统:
A Survey on Knowledge Graph-Based Recommender Systems (2020) 1
Graph Learning based Recommender Systems: A Review (2021) 2
近年来,将知识图谱(KG)作为补充信息引入推荐系统,引起了研究者的关注。KG是一个异构图,其中节点作为实体,边表示实体之间的关系。项目及其属性可以映射到KG中,以理解项目之间的相互关系。此外,用户和用户补充信息也可以集成到KG中,可以更准确地捕获用户和项目之间的关系以及用户偏好。
a. 信息网络(Information Network)
b. 异构信息网络(Heterogeneous Information Network, HIN)
c. 知识图谱(Knowledge Graph)
d. 元路径(Meta-path)
e. H-跳邻居(H-hop Neighbor)
f. 用户水波/涟漪集合(User Ripple Set)
交互矩阵可以表示成二部图(bipartite graph)。
序列交互可以表示成有向图(directed graph)
通常,我们使用三元组(head entity, relation, tail entity)来表示知识,例如(sky tree, location, Tokyo)。我们可以用独热向量来表示这个知识。但实体和关系太多,维度太大。当两个实体或关系很近时,独热向量无法捕捉相似度。受Wrod2Vec模型的启发,想用分布表示来表示实体和关系。
基于嵌入的方法通常直接使用来自KG的信息来丰富项目或用户的表示。为了利用KG信息,需要应用知识图嵌入(KGE)算法将KG编码为低维嵌入。
KG是由从数据集或外部知识库中提取的项目及其相关属性构建的,遵循该策略的论文利用知识图嵌入(KGE)算法对图进行编码,以便进行更全面的项目表示,然后将项目补充信息集成到推荐框架中。
(1) CKE12
(2) DKN13
DKN提出了一种深度知识感知网络,将知识图表示融入新闻推荐中。
DKN的关键组成部分是一个多渠道和词实体对齐的知识感知卷积神经网络(KCNN),它融合了新闻的语义级和知识级表示。
KCNN将单词和实体视为多个通道,并在卷积期间明确保持它们的对齐关系。此外,为了解决用户的各种兴趣,我们还在DKN中设计了一个注意力模块,以动态地聚合用户关于当前候选新闻的历史记录。
基于文本的知识图谱构建
首先,为了区分新闻内容中的知识实体,利用实体链接技术通过将它们与知识图中的预定义实体相关联来消除文本中的歧义提及。基于这些识别的实体,我们构建子图并从原始知识图中提取它们之间的所有关系链接。请注意,已识别实体之间的关系可能很稀疏且缺乏多样性。因此,我们将知识子图扩展到一跳内的所有实体。鉴于提取的知识图,许多知识图嵌入方法,如TransE,TransH,TransR和TransD,可用于实体表示学习。学习的实体嵌入被视为DKN框架中KCNN的输入。
(1) SHINE14
(2) KTUP15
(1) RippleNet16
Guo Q, Zhuang F, Qin C, et al. A survey on knowledge graph-based recommender systems[J]. IEEE Transactions on Knowledge and Data Engineering, 2020. ↩︎
Wang S, Hu L, Wang Y, et al. Graph learning based recommender systems: A review[J]. arXiv preprint arXiv:2105.06339, 2021. ↩︎
Wang X, Wang D, Xu C, et al. Explainable reasoning over knowledge graphs for recommendation[C]//Proceedings of the AAAI conference on artificial intelligence. 2019, 33(01): 5329-5336. ↩︎
Zheng L, Lu C T, Jiang F, et al. Spectral collaborative filtering[C]//Proceedings of the 12th ACM conference on recommender systems. 2018: 311-319. ↩︎
Qiu R, Li J, Huang Z, et al. Rethinking the item order in session-based recommendation with graph neural networks[C]//Proceedings of the 28th ACM international conference on information and knowledge management. 2019: 579-588. ↩︎
Qiu R, Li J, Huang Z, et al. Rethinking the item order in session-based recommendation with graph neural networks[C]//Proceedings of the 28th ACM international conference on information and knowledge management. 2019: 579-588. ↩︎
Fan W, Ma Y, Li Q, et al. Graph neural networks for social recommendation[C]//The world wide web conference. 2019: 417-426. ↩︎
Huang J, Ren Z, Zhao W X, et al. Taxonomy-aware multi-hop reasoning networks for sequential recommendation[C]//Proceedings of the Twelfth ACM International Conference on Web Search and Data Mining. 2019: 573-581. ↩︎
Bordes A, Usunier N, Garcia-Duran A, et al. Translating embeddings for modeling multi-relational data[J]. Advances in neural information processing systems, 2013, 26. ↩︎
Wang Z, Zhang J, Feng J, et al. Knowledge graph embedding by translating on hyperplanes[C]//Proceedings of the AAAI Conference on Artificial Intelligence. 2014, 28(1). ↩︎
Lin Y, Liu Z, Sun M, et al. Learning entity and relation embeddings for knowledge graph completion[C]//Twenty-ninth AAAI conference on artificial intelligence. 2015. ↩︎
Zhang F, Yuan N J, Lian D, et al. Collaborative knowledge base embedding for recommender systems[C]//Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining. 2016: 353-362. ↩︎
Wang H, Zhang F, Xie X, et al. DKN: Deep knowledge-aware network for news recommendation[C]//Proceedings of the 2018 world wide web conference. 2018: 1835-1844. ↩︎
Wang H, Zhang F, Hou M, et al. Shine: Signed heterogeneous information network embedding for sentiment link prediction[C]//Proceedings of the eleventh ACM international conference on web search and data mining. 2018: 592-600. ↩︎
Cao Y, Wang X, He X, et al. Unifying knowledge graph learning and recommendation: Towards a better understanding of user preferences[C]//The world wide web conference. 2019: 151-161. ↩︎
Wang H, Zhang F, Wang J, et al. Ripplenet: Propagating user preferences on the knowledge graph for recommender systems[C]//Proceedings of the 27th ACM international conference on information and knowledge management. 2018: 417-426. ↩︎
Copyright © 2003-2013 www.wpsshop.cn 版权所有,并保留所有权利。