Hgformer: Hyperbolic Graph Transformer for Collaborative Filtering

Xin Yang, Xingrun Li, Heng Chang, Yang Jinze, Xihong Yang, Shengyu Tao, Maiko Shigeno, Ningkang Chang, Junfeng Wang, Dawei Yin, Erxue Min
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:70813-70832, 2025.

Abstract

Recommender systems are increasingly spreading to different areas like e-commerce or video streaming to alleviate information overload. One of the most fundamental methods for recommendation is Collaborative Filtering (CF), which leverages historical user-item interactions to infer user preferences. In recent years, Graph Neural Networks (GNNs) have been extensively studied to capture graph structures in CF tasks. Despite this remarkable progress, local structure modeling and embedding distortion still remain two notable limitations in the majority of GNN-based CF methods. Therefore, in this paper, we propose a novel Hyperbolic Graph Transformer architecture, to tackle the long-tail problems in CF tasks. Specifically, the proposed framework is comprised of two essential modules: 1) Local Hyperbolic Graph Convolutional Network (LHGCN), which performs graph convolution entirely in the hyperbolic manifold and captures the local structure of each node; 2) Hyperbolic Transformer, which is comprised of hyperbolic cross-attention mechanisms to capture global information. Furthermore, to enable its feasibility on large-scale data, we introduce an unbiased approximation of the cross-attention for linear computational complexity, with a theoretical guarantee in approximation errors. Empirical experiments demonstrate that our proposed model outperforms the leading collaborative filtering methods and significantly mitigates the long-tail issue in CF tasks. Our implementations are available in https://github.com/EnkiXin/Hgformer.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-yang25o, title = {Hgformer: Hyperbolic Graph Transformer for Collaborative Filtering}, author = {Yang, Xin and Li, Xingrun and Chang, Heng and Jinze, Yang and Yang, Xihong and Tao, Shengyu and Shigeno, Maiko and Chang, Ningkang and Wang, Junfeng and Yin, Dawei and Min, Erxue}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {70813--70832}, year = {2025}, editor = {Singh, Aarti and Fazel, Maryam and Hsu, Daniel and Lacoste-Julien, Simon and Berkenkamp, Felix and Maharaj, Tegan and Wagstaff, Kiri and Zhu, Jerry}, volume = {267}, series = {Proceedings of Machine Learning Research}, month = {13--19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v267/main/assets/yang25o/yang25o.pdf}, url = {https://proceedings.mlr.press/v267/yang25o.html}, abstract = {Recommender systems are increasingly spreading to different areas like e-commerce or video streaming to alleviate information overload. One of the most fundamental methods for recommendation is Collaborative Filtering (CF), which leverages historical user-item interactions to infer user preferences. In recent years, Graph Neural Networks (GNNs) have been extensively studied to capture graph structures in CF tasks. Despite this remarkable progress, local structure modeling and embedding distortion still remain two notable limitations in the majority of GNN-based CF methods. Therefore, in this paper, we propose a novel Hyperbolic Graph Transformer architecture, to tackle the long-tail problems in CF tasks. Specifically, the proposed framework is comprised of two essential modules: 1) Local Hyperbolic Graph Convolutional Network (LHGCN), which performs graph convolution entirely in the hyperbolic manifold and captures the local structure of each node; 2) Hyperbolic Transformer, which is comprised of hyperbolic cross-attention mechanisms to capture global information. Furthermore, to enable its feasibility on large-scale data, we introduce an unbiased approximation of the cross-attention for linear computational complexity, with a theoretical guarantee in approximation errors. Empirical experiments demonstrate that our proposed model outperforms the leading collaborative filtering methods and significantly mitigates the long-tail issue in CF tasks. Our implementations are available in https://github.com/EnkiXin/Hgformer.} }
Endnote
%0 Conference Paper %T Hgformer: Hyperbolic Graph Transformer for Collaborative Filtering %A Xin Yang %A Xingrun Li %A Heng Chang %A Yang Jinze %A Xihong Yang %A Shengyu Tao %A Maiko Shigeno %A Ningkang Chang %A Junfeng Wang %A Dawei Yin %A Erxue Min %B Proceedings of the 42nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Aarti Singh %E Maryam Fazel %E Daniel Hsu %E Simon Lacoste-Julien %E Felix Berkenkamp %E Tegan Maharaj %E Kiri Wagstaff %E Jerry Zhu %F pmlr-v267-yang25o %I PMLR %P 70813--70832 %U https://proceedings.mlr.press/v267/yang25o.html %V 267 %X Recommender systems are increasingly spreading to different areas like e-commerce or video streaming to alleviate information overload. One of the most fundamental methods for recommendation is Collaborative Filtering (CF), which leverages historical user-item interactions to infer user preferences. In recent years, Graph Neural Networks (GNNs) have been extensively studied to capture graph structures in CF tasks. Despite this remarkable progress, local structure modeling and embedding distortion still remain two notable limitations in the majority of GNN-based CF methods. Therefore, in this paper, we propose a novel Hyperbolic Graph Transformer architecture, to tackle the long-tail problems in CF tasks. Specifically, the proposed framework is comprised of two essential modules: 1) Local Hyperbolic Graph Convolutional Network (LHGCN), which performs graph convolution entirely in the hyperbolic manifold and captures the local structure of each node; 2) Hyperbolic Transformer, which is comprised of hyperbolic cross-attention mechanisms to capture global information. Furthermore, to enable its feasibility on large-scale data, we introduce an unbiased approximation of the cross-attention for linear computational complexity, with a theoretical guarantee in approximation errors. Empirical experiments demonstrate that our proposed model outperforms the leading collaborative filtering methods and significantly mitigates the long-tail issue in CF tasks. Our implementations are available in https://github.com/EnkiXin/Hgformer.
APA
Yang, X., Li, X., Chang, H., Jinze, Y., Yang, X., Tao, S., Shigeno, M., Chang, N., Wang, J., Yin, D. & Min, E.. (2025). Hgformer: Hyperbolic Graph Transformer for Collaborative Filtering. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:70813-70832 Available from https://proceedings.mlr.press/v267/yang25o.html.

Related Material