We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations

DBLP - CS Bibliography


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Information Retrieval

Title: Edge-Enhanced Global Disentangled Graph Neural Network for Sequential Recommendation

Abstract: Sequential recommendation has been a widely popular topic of recommender systems. Existing works have contributed to enhancing the prediction ability of sequential recommendation systems based on various methods, such as recurrent networks and self-attention mechanisms. However, they fail to discover and distinguish various relationships between items, which could be underlying factors which motivate user behaviors. In this paper, we propose an Edge-Enhanced Global Disentangled Graph Neural Network (EGD-GNN) model to capture the relation information between items for global item representation and local user intention learning. At the global level, we build a global-link graph over all sequences to model item relationships. Then a channel-aware disentangled learning layer is designed to decompose edge information into different channels, which can be aggregated to represent the target item from its neighbors. At the local level, we apply a variational auto-encoder framework to learn user intention over the current sequence. We evaluate our proposed method on three real-world datasets. Experimental results show that our model can get a crucial improvement over state-of-the-art baselines and is able to distinguish item features.
Comments: 13 pages, 7 figures, 5 tables. Submitted to ICDE 2022
Subjects: Information Retrieval (cs.IR); Machine Learning (cs.LG)
Cite as: arXiv:2111.10539 [cs.IR]
  (or arXiv:2111.10539v2 [cs.IR] for this version)

Submission history

From: Yunyi Li [view email]
[v1] Sat, 20 Nov 2021 08:15:20 GMT (2824kb,D)
[v2] Tue, 23 Nov 2021 01:30:26 GMT (2826kb,D)

Link back to: arXiv, form interface, contact.