Heterogeneous Information Network Embedding for Recommendation

By Chuan Shi, Binbin Hu, Wayne Xin Zhao, Philip S. Yu. IEEE Transactions on Knowledge and Data Engineering, 2017

1,030 citations recorded by the source corpus.

Due to the flexibility in modelling data heterogeneity, heterogeneous information network (HIN) has been adopted to characterize complex and heterogeneous auxiliary data in recommender systems, called HIN based recommendation. It is challenging to develop effective methods for HIN based recommendation in both extraction and exploitation of the information from HINs. Most of HIN based recommendation methods rely on path based similarity, which cannot fully mine latent structure features of users and items. In this paper, we propose a novel heterogeneous network embedding based approach for HIN based recommendation, called HERec. To embed HINs, we design a meta-path based random walk strategy to generate meaningful node sequences for network embedding. The learned node embeddings are first transformed by a set of fusion functions, and subsequently integrated into an extended matrix factorization (MF) model. The extended MF model together with fusion functions are jointly optimized for the rating prediction task. Extensive experiments on three real-world datasets demonstrate the effectiveness of the HERec model. Moreover, we show the capability of the HERec model for the cold-start problem, and reveal that the transformed embedding information from HINs can improve the recommendation performance.

Subfields: Artificial Intelligence, Computer Networks and Communications, Information Systems

View Paper: https://www.semanticscholar.org/paper/378f0a62471ef232c7730d8a67717afa5104ab21

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. Chuan Shi
3. Philip S. Yu
4. Xiang Wang
5. Shirui Pan
6. Tat-seng Chua
7. Chengqi Zhang
8. Xiao Wang

In Computer Networks and Communications

1. Chuan Shi
2. Xiaoran Xu
3. Hanning Zhou
4. Yan Zhang
5. Yuping Zhang
6. Xiangnan He
7. Xiao Wang
8. Xiang Wang

In Information Systems

1. Xiangnan He
2. Xiang Wang
3. Tat-seng Chua
4. Xing Xie
5. Hongzhi Yin
6. Chuan Shi
7. Wayne Xin Zhao
8. Tong Chen

Citations in the corpus

• Graph Learning: A Survey (2021, methodology): “[114] proposed the HERec model, which utilizes the heterogeneous information network embedding for providing accurate recommendations in social networks.”
• Distilling Structured Knowledge into Embeddings for Explainable and Accurate Recommendation (2019, methodology): “tem popularity, BPR [29], WMF [14, 24] and VAE [18]) and the hybrid models that exploit both content and collaborative signals forrecommendation(LibFM[28],heterogeneousgraphembeddingbased model HERec [31] and the base model in our framework Proposedbase as introduced in Sect. 2.1). 3.2.3 Evaluation ”
• Graph Neural Networks in Recommender Systems: A Survey (2020, methodology): “Previous works preprocess KG by knowledge graph embedding methods to learn the embeddings of entities and relations, such as [8, 59, 75, 79] or design meta-paths to aggregate the neighbor information [47, 50, 64, 74].”
• Learning Intents behind Interactions with Knowledge Graph for Recommendation (2021, background): “KG stores the structured information of real-world facts, such as item attributes, taxonomy, or external commonsense knowledge, in the form of a heterogeneous graph or heterogeneous information network [28, 29].”
• Metapath-guided Heterogeneous Graph Neural Network for Intent Recommendation (2019, methodology): “As a general information modeling method, Heterogeneous Information Network (HIN) [18], consisting of multiple types of objects and links, has been widely applied to many data mining tasks [10, 17, 18].”
• Heterogeneous Graph Attention Network (2019, methodology): “g method which performs meta-path based random walk and utilizes skip-gram to embed the heterogeneous graphs. Here we test all the meta-paths for metapath2vec and report the best performance. • HERec [27]: A heterogeneous graph embedding method which designs a type constraint strategy to filter the ”
• Learning Graph Embedding With Adversarial Training Methods (2019, background): “2932096 an important role in various data-mining tasks, including classification [1], clustering [2], recommendation [3]–[5], and”
• AliGraph: A Comprehensive Graph Neural Network Platform (2019, methodology): “Metapath2Vec [10] and HERec [44] formalize meta-path based random-walks to construct the heterogeneous neighborhood of a node and then leverage skip-gram models to perform node embeddings.”
• MAGNN: Metapath Aggregated Graph Neural Network for Heterogeneous Graph Embedding (2020, methodology): “. (1) The model does not leverage node content features, so it rarely performs well on heterogeneous graphs with rich node content features (e.g., metapath2vec [9], ESim [22], HIN2vec [11], and HERec [23]). (2) The model discards all intermediate nodes along the metapath by only considering two end ”
• A Survey on Heterogeneous Graph Embedding: Methods, Techniques, Applications and Sources (2020, neutral): “To tackle the above issues, various heterogeneous graph embedding methods have been proposed [2], [8], [9], [15],”
• Where to Go Next: Modeling Long- and Short-Term User Preferences for Point-of-Interest Recommendation (2020, methodology): “More recently, deep learning based methods, such as embedding learning (Feng et al. 2015; Shi et al. 2018; Zhang et al. 2019), neural CF (He et al. 2017; Yin et al. 2017), deep latent factor model (Cheng et al. 2018), and metric learning (Tay, Anh Tuan, and Hui 2018) models, achieve promising…”
• Deep Representation Learning for Social Network Analysis (2019, neutral): “HERec (Shi et al., 2019) and metapath2vec++ (Dong et al.”
• Leveraging Meta-path based Context for Top- N Recommendation with A Neural Co-Attention Model (2018, methodology): “In addition, existing HIN based recommendation [26, 41] methods mainly focus on the task of rating prediction, while we may only have implicit feedback available in practice.”
• Representation Learning for Attributed Multiplex Heterogeneous Network (2019, methodology): “Heterogeneous Network (HEN) PTE [34] Multi Single / metapath2vec [7] HERec [31] Attributed HEN (AHEN) HNE [3] Multi Single Attributed”
• Representation Learning for Dynamic Graphs: A Survey (2019, methodology): “Shi et al. (2018) take a similar approach as metapath2vec but aim at learning node embeddings that are geared more towards improving recommendation performance.”
• A Survey on Knowledge Graph-Based Recommender Systems (2020, methodology): “Traditional path-based methods [3], [75], [76], [77], [78], [79], [82] generally integrate MF with extracted meta-paths in HINs.”
• Self-supervised Heterogeneous Graph Neural Network with Co-contrastive Learning (2021, methodology): “For GraphSAGE, GAE, Mp2vec, HERec and DGI, we test all the meta-paths for them and report the best performance.”