Decoupled Graph Convolution Network for Inferring Substitutable and Complementary Items

Liu, Yiding; Gu, Yun; Ding, Zhuoye; Gao, Junchao; Guo, Ziyi; Bao, Yongjun; Yan, Weipeng

doi:10.1145/3340531.3412695

Cited by 44 publications

(47 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We adjust the model so that PGPR can infer the product relationship. DecGCN models the substitutability and complementarity of products in separated embedding spaces [27]. To test the performance of the MFI model, we treat MFI as a supervised prediction model to infer relationship.…”

Section: Results and Analysismentioning

confidence: 99%

“…Further, PMSC adopts a novel loss function with relation constraints to distinguish between the substitutes and complements [9], and LVAE links two variational auto-encoders to learn latent features over product reviews [8]. SPEM considers both textual information and relational constraints [10], and DecGCN exploits the graph structure to learn product representations in different relationship spaces [11]. Nevertheless, all these methods suffer from two drawbacks.…”

Section: Product Relationship Inferencementioning

confidence: 99%

“…They incorporate product embedding with constraints to further promote the distinction between two relationships. DecGCN encodes the semantic representation of products through GCN on an product graph and model product substitutability and complementarity in two spaces [11].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Inferring Substitutable and Complementary Products with Knowledge-Aware Path Reasoning based on Dynamic Policy Network

Yang¹,

Ye²,

Wang³

et al. 2021

Preprint

View full text Add to dashboard Cite

Inferring the substitutable and complementary products for a given product is an essential and fundamental concern for the recommender system. To achieve this, existing approaches take advantage of the knowledge graphs to learn more evidences for inference, whereas they often suffer from invalid reasoning for lack of elegant decision making strategies. Therefore, we propose a novel Knowledge-Aware Path Reasoning (KAPR) model which leverages the dynamic policy network to make explicit reasoning over knowledge graphs, for inferring the substitutable and complementary relationships.Our contributions can be highlighted as three aspects. Firstly, we model this inference scenario as a Markov Decision Process in order to accomplish a knowledge-aware path reasoning over knowledge graphs. Secondly, we integrate both structured and unstructured knowledge to provide adequate evidences for making accurate decision-making. Thirdly, we evaluate our model on a series of real-world datasets, achieving competitive performance compared with state-of-the-art approaches. Our code is released on https://gitee.com/yangzijing flower/kapr/tree/master .

show abstract

Section: Results and Analysismentioning

confidence: 99%

Section: Product Relationship Inferencementioning

confidence: 99%

See 1 more Smart Citation

Inferring Substitutable and Complementary Products with Knowledge-Aware Path Reasoning based on Dynamic Policy Network

Yang¹,

Ye²,

Wang³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…In this way, the whole structures to compute 𝑧 𝑣 𝑎 and 𝑧 𝑣 𝑞 are symmetric, which is also the reason we name this network Siamese neighbor matching layer. It's important to note that Siamese neighborhood matching mechanism differs from the dual matching scheme used in [14,16,31]. These studies design symmetric network structures for heterogeneous objects in two-tower; however, their neighborhood embedding is computed based on different metapaths and network parameters thus tend to be hard for heterogeneous matching.…”

Section: Siamese Neighbor Matchingmentioning

confidence: 99%

Heterogeneous Graph Neural Networks for Large-Scale Bid Keyword Matching

Liu

et al. 2021

Proceedings of the 30th ACM International Conference on Information &Amp; Knowledge Management

View full text Add to dashboard Cite

Digital advertising is a critical part of many e-commerce platforms such as Taobao and Amazon. While in recent years a lot of attention has been drawn to the consumer side including canonical problems like ctr/cvr prediction, the advertiser side, which directly serves advertisers by providing them with marketing tools, is now playing a more and more important role. When speaking of sponsored search, bid keyword recommendation is the fundamental service. This paper addresses the problem of keyword matching, the primary step of keyword recommendation. Existing methods for keyword matching merely consider modeling relevance based on a single type of relation among ads and keywords, such as query clicks or text similarity, which neglects rich heterogeneous interactions hidden behind them. To fill this gap, the keyword matching problem faces several challenges including: 1) how to learn enriched and robust embeddings from complex interactions among various types of objects; 2) how to conduct high-quality matching for new ads that usually lack sufficient data.To address these challenges, we develop a heterogeneous-graphneural-network-based model for keyword matching named Het-Match, which has been deployed both online and offline at the core sponsored search platform of Alibaba Group. To extract enriched and robust embeddings among rich relations, we design a hierarchical structure to fuse and enhance the relevant neighborhood patterns both on the micro and the macro level. Moreover, by proposing a multi-view framework, the model is able to involve more positive samples for cold-start ads. Experimental results on a large-scale industrial dataset as well as online AB tests exhibit the effectiveness of HetMatch. CCS CONCEPTS• Information systems → Retrieval models and ranking; Computational advertising; • Computing methodologies → Neural networks.

show abstract

“…Click-through rate (CTR) prediction, which is one of the critical task for ranking in Search Engines and Recommender Systems, has received much attention from both the research and industry community [4,12,14,21,29,37]. Deep learning based methods [4,5,9,10,28,36,37] have achieved state-of-the-art performance in CTR prediction by modeling users' sequential behaviors.…”

mentioning

confidence: 99%

Modeling Users' Contextualized Page-wise Feedback for Click-Through Rate Prediction in E-commerce Search

Fan

et al. 2022

Proceedings of the Fifteenth ACM International Conference on Web Search and Data Mining

Self Cite

View full text Add to dashboard Cite

Modeling user's historical feedback is essential for Click-Through Rate Prediction in personalized search and recommendation. Existing methods usually only model users' positive feedback information such as click sequences which neglects the context information of the feedback. In this paper, we propose a new perspective for context-aware users' behavior modeling by including the whole page-wisely exposed products and the corresponding feedback as contextualized page-wise feedback sequence. The intra-page context information and inter-page interest evolution can be captured to learn more specific user preference. We design a novel neural ranking model RACP(Recurrent Attention over Contextualized Page sequence), which utilizes page-context aware attention to model the intra-page context. A recurrent attention process is used to model the cross-page interest convergence evolution as denoising the interest in the previous pages. Experiments on public and real-world industrial datasets verify our model's effectiveness.

show abstract

Decoupled Graph Convolution Network for Inferring Substitutable and Complementary Items

Cited by 44 publications

References 21 publications

Inferring Substitutable and Complementary Products with Knowledge-Aware Path Reasoning based on Dynamic Policy Network

Inferring Substitutable and Complementary Products with Knowledge-Aware Path Reasoning based on Dynamic Policy Network

Heterogeneous Graph Neural Networks for Large-Scale Bid Keyword Matching

Modeling Users' Contextualized Page-wise Feedback for Click-Through Rate Prediction in E-commerce Search

Contact Info

Product

Resources

About