UltraGCN

Mao, Kelong; Zhu, Jieming; Xiao, Xi; Lu, Biao; Wang, Zhaowei; He, Xiuqiang

doi:10.1145/3459637.3482291

Cited by 203 publications

(36 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Each user and item has an embedding vector, and the dot product is used to represent the interaction possibility of different users and items. Including LR-GCCF [3], NIA-GCN [29], LightGCN [11], DGCF [30], NGAT4Rec [28], SGL-ED [33], UltraGCN [20] and DGCF [16]. • GSP-based GF-CF method [27], which sums up different collaborative filtering methods, including those based on the neighborhood, matrix factorization and graph neural network, into low-pass filters with different frequency response functions.…”

Section: Baselinesmentioning

confidence: 99%

Personalized Graph Signal Processing for Collaborative Filtering

Liu,

Li,

et al. 2023

Preprint

View full text Add to dashboard Cite

The collaborative filtering (CF) problem with only user-item interaction information can be solved by graph signal processing (GSP), which uses low-pass filters to smooth the observed interaction signals on the similarity graph to obtain the prediction signals.However, the interaction signal may not be sufficient to accurately characterize user interests and the low-pass filters may ignore the useful information contained in the high-frequency component of the observed signals, resulting in suboptimal accuracy. To this end, we propose a personalized graph signal processing (PGSP) method for collaborative filtering. Firstly, we design the personalized graph signal containing richer user information and construct an augmented similarity graph containing more graph topology information, to more effectively characterize user interests. Secondly, we devise a mixed-frequency graph filter to introduce useful information in the high-frequency components of the observed signals by combining an ideal low-pass filter that smooths signals globally and a linear low-pass filter that smooths signals locally. Finally, we combine the personalized graph signal, the augmented similarity graph and the mixed-frequency graph filter by proposing a pipeline consisting of three key steps: pre-processing, graph convolution and post-processing. Extensive experiments show that PGSP can achieve superior accuracy compared with state-of-the-art

show abstract

Section: Baselinesmentioning

confidence: 99%

Personalized Graph Signal Processing for Collaborative Filtering

Liu,

Li,

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…To alleviate the impact of spurious correlation, it sampled subgraphs to learn the invariant masks which divided the multimodal representation into invariant and variant representations. Based on the UltraGCN [35], utilizing the invariant representations could eliminate the spurious correlation.…”

Section: Heterogeneous Graph Fusionmentioning

confidence: 99%

A Comprehensive Survey on Multimodal Recommender Systems: Taxonomy, Evaluation, and Future Directions

Zhou¹,

Zhou²,

Zeng³

et al. 2023

Preprint

View full text Add to dashboard Cite

Recommendation systems have become popular and effective tools to help users discover their interesting items by modeling the user preference and item property based on implicit interactions (e.g., purchasing and clicking). Humans perceive the world by processing the modality signals (e.g., audio, text and image), which inspired researchers to build a recommender system that can understand and interpret data from different modalities. Those models could capture the hidden relations between different modalities and possibly recover the complementary information which can not be captured by a uni-modal approach and implicit interactions. The goal of this survey is to provide a comprehensive review of the recent research efforts on the multimodal recommendation. Specifically, it shows a clear pipeline with commonly used techniques in each step and classifies the models by the methods used. Additionally, a code framework has been designed that helps researchers new in this area to understand the principles and techniques, and easily runs the SOTA models. Our framework is located at: https://github.com/enoche/MMRec.

show abstract

“…Initialization. Following existing GCN-based recommendation models [9,11,15,32], we initialize the embedding vectors of a user 𝑢 ∈ U and an item 𝑖 ∈ I as 𝒆 0 𝒖 ∈ R 𝑑 and 𝒆 0 𝒊 ∈ R 𝑑 , respectively. 𝑑 denotes the embedding size.…”

Section: Embeddingmentioning

confidence: 99%

“…The model-based collaborative filtering, which learns user and item representations based on useritem interactions for recommendation, has become the mainstream recommendation technique since its success in the Netflix contest [12]. With the rise of deep learning, CF methods have gained rapid development from shallow models [3,12,18] to deep models [4,10,15,24]. The deep neural network (DNN) based models can learn better user/item representations and capture more complicated user-item interaction relations [2,9,27,35].…”

Section: Introductionmentioning

confidence: 99%

“…The deep neural network (DNN) based models can learn better user/item representations and capture more complicated user-item interaction relations [2,9,27,35]. And the graph convolutional network (GCN) based ones exploit the high-order proximities over the user-item interaction graph to enhance the representation learning [9,13,15,24]. In addition, the DNN-and GCN-based models are often equipped with advanced techniques, such as the attention mechanism [23], disentangled learning [25], contrastive learning [26,34], setting up a new standard of recommendation.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Multi-Behavior Recommendation with Cascading Graph Convolution Networks

Cheng

Han

Liu

et al. 2023

Proceedings of the ACM Web Conference 2023

View full text Add to dashboard Cite

Multi-behavior recommendation, which exploits auxiliary behaviors (e.g., click and cart) to help predict users' potential interactions on the target behavior (e.g., buy), is regarded as an effective way to alleviate the data sparsity or cold-start issues in recommendation. Multi-behaviors are often taken in certain orders in real-world applications (e.g., click>cart>buy). In a behavior chain, a latter behavior usually exhibits a stronger signal of user preference than the former one does. Most existing multi-behavior models fail to capture such dependencies in a behavior chain for embedding learning. In this work, we propose a novel multi-behavior recommendation model with cascading graph convolution networks (named MB-CGCN). In MB-CGCN, the embeddings learned from one behavior are used as the input features for the next behavior's embedding learning after a feature transformation operation. In this way, our model explicitly utilizes the behavior dependencies in embedding learning. Experiments on two benchmark datasets demonstrate the effectiveness of our model on exploiting multi-behavior data. It outperforms the best baseline by 33.7% and 35.9% on average over the two datasets in terms of Recall@10 and NDCG@10, respectively.

show abstract

UltraGCN

Cited by 203 publications

References 24 publications

Personalized Graph Signal Processing for Collaborative Filtering

Personalized Graph Signal Processing for Collaborative Filtering

A Comprehensive Survey on Multimodal Recommender Systems: Taxonomy, Evaluation, and Future Directions

Multi-Behavior Recommendation with Cascading Graph Convolution Networks

Contact Info

Product

Resources

About