Multi-Facet Recommender Networks with Spherical Optimization

Tan, Yanchao; Yang, Carl; Wei, Xiangyu; Ma, Yue; Zheng, Xiaolin

doi:10.1109/icde51399.2021.00135

Cited by 12 publications

(6 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…toward trivial solutions, especially with difficult samples [7]. Such "lazy learning" with localized similarities can inadvertently stifle the learning of diversity, which may bias the system towards common or popular preferences, thus suppressing the variety inherent in user behavior.…”

Section: Profiling Users With Both Interactions and Tagsmentioning

confidence: 99%

Profiling Users with Tag-enhanced Spherical Metric Learning for Recommendation

Tan,

Lv,

Huang

et al. 2024

Preprint

View full text Add to dashboard Cite

With the growing complex interactions between users and items on the Internet, the difficulty of representing them has increased in Recommender Systems (RSs). Existing methods propose to capture the complex user-item interactions that implicitly associate a user with other like-minded ones. However, the learned embeddings cannot efficiently capture the similar patterns underlying users’ behaviors. Moreover, the implicitly retrieved patterns cannot be readily accurate and interpretable. To this end, we propose to leverage both users' interacted items and item tags, so as to advocate for tag-enhanced recommendations. Specifically, we propose an adaptive tag selection mechanism to ensure the quality of tag representations, so as to profile users and learn tag-enhanced representations of users/items. Additionally, we introduce a spherical optimization strategy to fully exploit the increased representation capacity of the tag-enhanced recommendation framework. We conduct numerous experiments on four real-world datasets, where our proposed tag-enhanced framework for recommendation can bring consistent performance gains and averagely achieve a 12.54% improvement regarding both Recall and NDCG metrics.

show abstract

Section: Profiling Users With Both Interactions and Tagsmentioning

confidence: 99%

Profiling Users with Tag-enhanced Spherical Metric Learning for Recommendation

Tan,

Lv,

Huang

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…We use grid search to find the optimal hyper-parameters for HybridGNN. Specifically, we set the basic embedding e vi as [64, 128, 256, 512], the local edge embedding e vi,r l as [2, 8, 64, 128] and the negative node numbers as [1,3,5,7]. Moreover, the batch size is 2048 for all the datasets.…”

Section: Experimental Settingsmentioning

confidence: 99%

“…In recent years, exploiting user interests through network embedding has emerged as a trending research topic in the field of recommender systems [1]- [3]. Traditional graph embedding methods such as DeepWalk [4], node2vec [5] and graph neural networks like GCN [6], GraphSage [7] have achieved good performances in many tasks such as node Fig.…”

Section: Introductionmentioning

confidence: 99%

HybridGNN: Learning Hybrid Representation for Recommendation in Multiplex Heterogeneous Networks

Wang

et al. 2022

2022 IEEE 38th International Conference on Data Engineering (ICDE)

View full text Add to dashboard Cite

Recently, graph neural networks have shown the superiority of modeling the complex topological structures in heterogeneous network-based recommender systems. Due to the diverse interactions among nodes and abundant semantics emerging from diverse types of nodes and edges, there is a bursting research interest in learning expressive node representations in multiplex heterogeneous networks. One of the most important tasks in recommender systems is to predict the potential connection between two nodes under a specific edge type (i.e., relationship). Although existing studies utilize explicit metapaths to aggregate neighbors, practically they only consider intra-relationship metapaths and thus fail to leverage the potential uplift by inter-relationship information. Moreover, it is not always straightforward to exploit inter-relationship metapaths comprehensively under diverse relationships, especially with the increasing number of node and edge types. In addition, contributions of different relationships between two nodes are difficult to measure. To address the challenges, we propose HybridGNN, an end-to-end GNN model with hybrid aggregation flows and hierarchical attentions to fully utilize the heterogeneity in the multiplex scenarios. Specifically, HybridGNN applies a randomized inter-relationship exploration module to exploit the multiplexity property among different relationships. Then, our model leverages hybrid aggregation flows under intrarelationship metapaths and randomized exploration to learn the rich semantics. To explore the importance of different aggregation flow and take advantage of the multiplexity property, we bring forward a novel hierarchical attention module which leverages both metapath-level attention and relationship-level attention. Extensive experimental results on five real-world datasets suggest that HybridGNN achieves the best performance compared to several state-of-the-art baselines (p < 0.01, t-test) with statistical significance.

show abstract

“…How to comprehensively utilize the cross domain information to improve the performance of recommendation systems has become a hot topic. Therefore, Cross Domain Recommendation (CDR) becomes more and more attractive for establishing highly accurate recommendation systems [5,25,43,54]. Most existing CDR models assume that both the source and target domains share the same set of users, which makes it easier to transfer useful knowledge across domains based on these overlapped users.…”

Section: Introductionmentioning

confidence: 99%

Collaborative Filtering with Attribution Alignment for Review-based Non-overlapped Cross Domain Recommendation

Liu¹,

Zheng²,

Hu³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Cross-Domain Recommendation (CDR) has been popularly studied to utilize different domain knowledge to solve the data sparsity and cold-start problem in recommender systems. In this paper, we focus on the Review-based Non-overlapped Recommendation (RNCDR) problem. The problem is commonly-existed and challenging due to two main aspects, i.e, there are only positive user-item ratings on the target domain and there is no overlapped user across different domains. Most previous CDR approaches cannot solve the RNCDR problem well, since (1) they cannot effectively combine review with other information (e.g., ID or ratings) to obtain expressive user or item embedding, (2) they cannot reduce the domain discrepancy on users and items. To fill this gap, we propose Collaborative Filtering with Attribution Alignment model (CFAA), a cross-domain recommendation framework for the RNCDR problem. CFAA includes two main modules, i.e., rating prediction module and embedding attribution alignment module. The former aims to jointly mine review, one-hot ID, and multi-hot historical ratings to generate expressive user and item embeddings. The later includes vertical attribution alignment and horizontal attribution alignment, tending to reduce the discrepancy based on multiple perspectives. Our empirical study on Douban and Amazon datasets demonstrates that CFAA significantly outperforms the state-of-the-art models under the RNCDR setting.

show abstract

Multi-Facet Recommender Networks with Spherical Optimization

Cited by 12 publications

References 40 publications

Profiling Users with Tag-enhanced Spherical Metric Learning for Recommendation

Profiling Users with Tag-enhanced Spherical Metric Learning for Recommendation

HybridGNN: Learning Hybrid Representation for Recommendation in Multiplex Heterogeneous Networks

Collaborative Filtering with Attribution Alignment for Review-based Non-overlapped Cross Domain Recommendation

Contact Info

Product

Resources

About