KGAT: Knowledge Graph Attention Network for Recommendation

By Xiang Wang, Xiangnan He, Yixin Cao, Meng Liu, Tat-seng Chua. Knowledge Discovery and Data Mining, 2019

2,302 citations recorded by the source corpus.

To provide more accurate, diverse, and explainable recommendation, it is compulsory to go beyond modeling user-item interactions and take side information into account. Traditional methods like factorization machine (FM) cast it as a supervised learning problem, which assumes each interaction as an independent instance with side information encoded. Due to the overlook of the relations among instances or items (e.g., the director of a movie is also an actor of another movie), these methods are insufficient to distill the collaborative signal from the collective behaviors of users. In this work, we investigate the utility of knowledge graph (KG), which breaks down the independent interaction assumption by linking items with their attributes. We argue that in such a hybrid structure of KG and user-item graph, high-order relations --- which connect two items with one or multiple linked attributes --- are an essential factor for successful recommendation. We propose a new method named Knowledge Graph Attention Network (KGAT) which explicitly models the high-order connectivities in KG in an end-to-end fashion. It recursively propagates the embeddings from a node's neighbors (which can be users, items, or attributes) to refine the node's embedding, and employs an attention mechanism to discriminate the importance of the neighbors. Our KGAT is conceptually advantageous to existing KG-based recommendation methods, which either exploit high-order relations by extracting paths or implicitly modeling them with regularization. Empirical results on three public benchmarks show that KGAT significantly outperforms state-of-the-art methods like Neural FM and RippleNet. Further studies verify the efficacy of embedding propagation for high-order relation modeling and the interpretability benefits brought by the attention mechanism. We release the codes and datasets at https://github.com/xiangwang1223/knowledge_graph_attention_network.

Subfields: Artificial Intelligence, Information Systems, Molecular Biology

View Paper: https://www.semanticscholar.org/paper/de02ec03f6a71246e505862a7195894601fbab99

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Xiangnan He
2. Xiang Wang
3. Tat-seng Chua
4. Liqiang Nie
5. Fuli Feng
6. Yong Li
7. Chen Gao
8. Xing Xie

In Information Systems

1. Xiangnan He
2. Xiang Wang
3. Tat-seng Chua
4. Fuli Feng
5. Liqiang Nie
6. Xing Xie
7. Meng Wang
8. Yong Li

In Molecular Biology

1. Xiangnan He
2. Xiang Wang
3. Tat-seng Chua
4. Zhiyong Cheng
5. Liqiang Nie
6. Xing Xie
7. Chuan Shi
8. Chang Zhou

Citations in the corpus

• Learning Intents behind Interactions with Knowledge Graph for Recommendation (2021, methodology): “Experimental results on three benchmark datasets show that, KGIN achieves significant improvements over the state-ofthe-art methods like KGAT [41], KGNN-LS [38], and CKAN [47].”
• Sequential Recommendation with Graph Neural Networks (2021, methodology): “GCN-based methods have achieved the state-of-the-art performance in other recommendation problems, such as social recommendation [6, 32, 40], knowledge graph-based recommendation [27, 28], multi-behavior recommendation [12], bundle recommendation [3], etc.”
• Self-supervised Graph Learning for Recommendation (2020, methodology): “Recently, attention mechanism is introduced into GCN-based recommendation models [43], which”
• A federated graph neural network framework for privacy-preserving personalization (2021, methodology): “For example, Wang et al. [30] proposed a knowledge-graph enhanced recommendation approach based on Knowledge Graph Attention Network (KGAT).”
• A Comprehensive Survey on Automatic Knowledge Graph Construction (2023, background): “1 INTRODUCTION Knowledge graphs (KGs) provide well-organized human knowledge for applications like search engines [1], recommendation systems [2], and question answering [3].”
• A Survey of Graph Neural Networks for Recommender Systems: Challenges, Methods, and Directions (2021, methodology): “Model Network Design Source of Explainability ECFKG [5] Meta-path based GE Meta-path over knowledge graph RippleNet [171] Meta-path based GraphSage Meta-path over knowledge graph EIUM [75] Meta-path based GE Meta-path over knowledge graph KPRN [185] Meta-path based GE Meta-path over knowledge graph ”
• Disentangled Graph Collaborative Filtering (2020, methodology): “history with users [21], item knowledge [15, 32, 39, 41, 42], and user reviews [5, 22], or conduct psychology experiments [24] to establish the ground truth on user intents and further interpret disentangled representations better.”
• A Survey on Knowledge Graph-Based Recommender Systems (2020, methodology): “Book-Crossing, DB-book2014, IntentBooks, and Amazon-Book contain binary feedback between users and books, and the KG for each dataset is built by mapping books to corresponding entities in Satori [2], [14], [45], [88], [89], [91], [92], [93], [95], DBpe-dia [70], [87], or Freebase [44], [90], [93].”
• Are we really making much progress?: Revisiting, benchmarking and refining heterogeneous graph neural networks (2021, methodology): “Following KGAT [35], we opt for BPR [27] loss for training.”
• LightGCN: Simplifying and Powering Graph Convolution Network for Recommendation (2020, background): “For example, a recent trend is to exploit auxiliary information such as item knowledge graph [38], social network [41] and multimedia content [44] for recommendation, where GCNs have set up the new state-of-the-art.”
• LARA: Attribute-to-feature Adversarial Learning for New-item Recommendation (2020, background): “Aiming at providing a ranking list of items to a specific user, a personalized recommender calculates the predicted relevance between users and items [2, 21, 35].”
• Graph Transformer Networks (2019, background): “Also, as several heterogeneous graph datasets have been recently studied for other network analysis tasks, such as link prediction [36, 41] and graph classification [17, 24], applying our GTNs to the other tasks can be interesting future directions.”
• Neural Graph Collaborative Filtering (2019, methodology): “Specifically, there are many other forms of structural information can be useful for understanding user behaviors, such as the cross features [40] in context-aware and semantics-rich recommendation [22, 27], item knowledge graph [31], and social networks [33].”
• Meta-learning on Heterogeneous Information Networks for Cold-start Recommendation (2020, methodology): “Some methods also integrate review texts, images or knowledge graphs to further enhance user and item representations [30, 34].”