pith. sign in

arxiv: 2504.13614 · v2 · submitted 2025-04-18 · 💻 cs.IR · cs.AI· cs.NE

Adaptive Long-term Embedding with Denoising and Augmentation for Recommendation

Zahra Akhlaghi , Mostafa Haghir Chehreghani This is my paper

Pith reviewed 2026-05-22 19:17 UTC · model grok-4.3

classification 💻 cs.IR cs.AIcs.NE
keywords recommendation systemsgraph neural networksdenoisingcommunity detectionlong-term embeddingsadaptive weightingsequential recommendation
0
0 comments X

The pith

ALDA4Rec improves recommendation accuracy by denoising item-item graphs with community detection and adaptively weighting long-term embeddings.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper tries to establish that a new method called ALDA4Rec can provide more accurate and robust recommendations than current approaches. It does so by building an item-item graph from interactions, applying community detection to filter noise, and then using GCNs for short-term user representations along with GRUs, attention, and adaptive MLP weighting for long-term preferences. This matters because recommendation systems often struggle with noisy data and fail to capture evolving user interests over time. If successful, such a model would lead to better personalized suggestions in online platforms.

Core claim

ALDA4Rec constructs an item-item graph, filters noise through community detection, and enriches user-item interactions. Graph Convolutional Networks (GCNs) are then employed to learn short-term representations, while averaging, GRUs, and attention mechanisms are utilized to model long-term embeddings. An MLP-based adaptive weighting strategy is further incorporated to dynamically optimize long-term user preferences. Experiments conducted on four real-world datasets demonstrate that ALDA4Rec outperforms state-of-the-art baselines, delivering notable improvements in both accuracy and robustness.

What carries the argument

ALDA4Rec's pipeline of item-item graph construction with community detection denoising, GCN-based short-term learning, and MLP-adaptive fusion of long-term embeddings from GRUs and attention.

If this is right

  • Outperforms state-of-the-art methods in accuracy on four real-world datasets.
  • Provides more robust recommendations in the presence of noise.
  • Dynamically optimizes long-term user preferences using adaptive weighting.
  • Captures complex user-item interactions through graph-based representations.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The approach of using community detection for denoising could be extended to other types of graph-based models in information retrieval.
  • Adaptive mechanisms for long-term embeddings may find use in related areas like user behavior prediction.
  • The overall framework suggests potential for hybrid models that combine denoising with sequential modeling techniques.

Load-bearing premise

Community detection on the constructed item-item graph reliably separates noise from useful signals without discarding important user-item interactions that affect downstream embedding quality.

What would settle it

If the performance improvements disappear when community detection is not used or when it is replaced by a different noise filtering method on the four datasets, the central role of that step would be disproven.

Figures

Figures reproduced from arXiv: 2504.13614 by Mostafa Haghir Chehreghani, Zahra Akhlaghi.

Figure 1
Figure 1. Figure 1: Overall framework of the proposed ALDA4Rec model [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Extracting user interactions from I sim t Finally, we identify items that do not belong to any community as noise since the user is not interested in their similar items, and they remain ungrouped. We then reduce their associated edge weights in the matrix At , with the degree of reduction determined by a hyperparameter. This ad￾justment minimizes the influence of noisy items on subsequent analyses, enhanc… view at source ↗
Figure 3
Figure 3. Figure 3: Performance of our model for various values of [PITH_FULL_IMAGE:figures/full_fig_p019_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Analysis of the Performance of Our Model for Various Values of [PITH_FULL_IMAGE:figures/full_fig_p019_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Impact of minsim variation on model performance in HR@10. 6 Conclusion In this paper, we proposed a sequential recommendation system leveraging GNNs to enhance recom￾mendation accuracy. Our approach offers two primary contributions. First, we introduced a novel graph construction and denoising technique that refines user interaction data by reducing noise and 19 [PITH_FULL_IMAGE:figures/full_fig_p019_5.png] view at source ↗
read the original abstract

The rapid growth of the internet has made personalized recommendation systems indispensable. Graph-based sequential recommendation systems, powered by Graph Neural Networks (GNNs), effectively capture complex user-item interactions but often face challenges such as noise and static representations. In this paper, we introduce the Adaptive Long-term Embedding with Denoising and Augmentation for Recommendation (ALDA4Rec) method, a novel model that constructs an item-item graph, filters noise through community detection, and enriches user-item interactions. Graph Convolutional Networks (GCNs) are then employed to learn short-term representations, while averaging, GRUs, and attention mechanisms are utilized to model long-term embeddings. An MLP-based adaptive weighting strategy is further incorporated to dynamically optimize long-term user preferences. Experiments conducted on four real-world datasets demonstrate that ALDA4Rec outperforms state-of-the-art baselines, delivering notable improvements in both accuracy and robustness. The source code is available at https://github.com/zahraakhlaghi/ALDA4Rec.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The paper proposes ALDA4Rec, a sequential recommendation model that first builds an item-item graph from user-item data, applies community detection to denoise the graph, employs GCNs to learn short-term embeddings, models long-term user preferences using averaging combined with GRUs and attention, and uses an MLP to adaptively weight the long-term embeddings. Experiments on four real-world datasets are reported to show outperformance over state-of-the-art baselines in accuracy and robustness, with source code released.

Significance. If the performance gains hold after proper validation, the work offers a practical pipeline that integrates graph denoising with adaptive long-term modeling, which could improve robustness in noisy recommendation settings. The release of source code supports reproducibility and is a clear strength.

major comments (3)
  1. [Experiments] The experimental section (results tables): no error bars, standard deviations, statistical significance tests, or details on data splits and hyperparameter tuning are provided, despite the abstract claiming consistent outperformance on four datasets. This undermines verification of the central accuracy and robustness claims.
  2. [Methodology] Methodology section on graph construction and community detection: no ablation isolating the denoising step (e.g., performance with vs. without community detection, or edge retention statistics) is reported. Without this, it remains unclear whether the detected communities preferentially remove noise while preserving predictive user-item paths, which is load-bearing for attributing gains to ALDA4Rec rather than the GCN + GRU/attention components.
  3. [Methodology] The MLP adaptive weighting subsection: the coefficients are described as dynamically optimizing long-term preferences, but no details on training objective, regularization, or sensitivity analysis are given, leaving open the possibility that gains arise from additional fitting capacity rather than the claimed adaptivity.
minor comments (2)
  1. [Abstract] The abstract states that the model 'enriches user-item interactions' but does not specify the augmentation technique; this should be clarified with a brief description or reference to the relevant subsection.
  2. Notation for short-term vs. long-term embeddings should be introduced consistently with explicit symbols (e.g., e_s for short-term) to improve readability across sections.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for their thorough review and valuable suggestions. We will revise the manuscript to address the concerns regarding experimental validation and methodological details, thereby strengthening the presentation of our work.

read point-by-point responses

  1. Referee: [Experiments] The experimental section (results tables): no error bars, standard deviations, statistical significance tests, or details on data splits and hyperparameter tuning are provided, despite the abstract claiming consistent outperformance on four datasets. This undermines verification of the central accuracy and robustness claims.

    Authors: We agree that providing error bars, standard deviations, and statistical significance tests is essential for robust claims. In the revised manuscript, we will report results with error bars from multiple random seeds, include standard deviations, perform statistical tests such as Wilcoxon signed-rank tests to compare with baselines, and detail the data splitting strategy (e.g., leave-one-out or temporal split) along with the hyperparameter search process. These additions will allow better verification of the reported improvements. revision: yes

  2. Referee: [Methodology] Methodology section on graph construction and community detection: no ablation isolating the denoising step (e.g., performance with vs. without community detection, or edge retention statistics) is reported. Without this, it remains unclear whether the detected communities preferentially remove noise while preserving predictive user-item paths, which is load-bearing for attributing gains to ALDA4Rec rather than the GCN + GRU/attention components.

    Authors: We acknowledge the importance of isolating the contribution of the community detection-based denoising. We will include an ablation study in the revised version, presenting performance metrics with and without the denoising step. Additionally, we will report statistics on the number of edges retained after community detection to illustrate the noise removal process and its impact on preserving relevant interactions. revision: yes

  3. Referee: [Methodology] The MLP adaptive weighting subsection: the coefficients are described as dynamically optimizing long-term preferences, but no details on training objective, regularization, or sensitivity analysis are given, leaving open the possibility that gains arise from additional fitting capacity rather than the claimed adaptivity.

    Authors: We will expand the description of the MLP adaptive weighting in the revised manuscript. This will include the specific training objective (e.g., the recommendation loss combined with any auxiliary losses), regularization methods employed (such as dropout or L2), and a sensitivity analysis varying the MLP architecture and hyperparameters to demonstrate the robustness and adaptive nature of the weighting strategy. revision: yes

Circularity Check

0 steps flagged

No circularity in derivation; model is standard component composition

full rationale

The paper presents ALDA4Rec as a pipeline of graph construction, community detection denoising, GCN short-term embeddings, GRU/attention long-term modeling, and MLP adaptive weighting. No equations, first-principles derivations, or predictions appear that reduce to fitted inputs or self-citations by construction. Performance claims rest on external dataset experiments rather than internal tautologies, rendering the approach self-contained against benchmarks with no load-bearing circular steps.

Axiom & Free-Parameter Ledger

2 free parameters · 1 axioms · 0 invented entities

The approach rests on standard graph neural network assumptions plus the domain-specific premise that community detection removes noise without harming signal; several hyperparameters for graph construction, community detection, and the MLP are expected to be tuned on data.

free parameters (2)
  • community detection resolution or threshold
    Parameter controlling how aggressively noise is filtered from the item-item graph.
  • MLP adaptive weighting coefficients
    Learned or tuned weights that balance short-term and long-term embeddings.
axioms (1)
  • domain assumption Community detection on user-item derived graphs separates noise from meaningful item co-occurrences
    Invoked when the paper states it filters noise through community detection.

pith-pipeline@v0.9.0 · 5707 in / 1339 out tokens · 46450 ms · 2026-05-22T19:17:18.425214+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

43 extracted references · 43 canonical work pages · 1 internal anchor

  1. [1]

    Federico Barbero, Andrea Banino, Steven Kapturowski, Dharshan Kumaran, Jo˜ ao G. M. Ara´ ujo, Alex Vitvitskyi, Razvan Pascanu, and Petar Velickovic. Transformers need glasses! information over-squashing in language tasks. CoRR, abs/2406.04267, 2024

    work page arXiv 2024

  2. [2]

    Best of both worlds: Advantages of hybrid graph sequence models

    Ali Behrouz, Ali Parviz, Mahdi Karami, Clayton Sanford, Bryan Perozzi, and Vahab Mirrokni. Best of both worlds: Advantages of hybrid graph sequence models. CoRR, abs/2411.15671, 2024

    work page arXiv 2024

  3. [3]

    Blondel, Jean-Loup Guillaume, and Renaud Lambiotte

    Vincent D. Blondel, Jean-Loup Guillaume, and Renaud Lambiotte. Fast unfolding of communi- ties in large networks: 15 years later. CoRR, abs/2311.06047, 2023

    work page arXiv 2023

  4. [4]

    Half a decade of graph convolutional networks

    Mostafa Haghir Chehreghani. Half a decade of graph convolutional networks. Nat. Mach. Intell. , 4(3):192–193, 2022

    work page 2022

  5. [5]

    Revisiting graph based collab- orative filtering: A linear residual graph convolutional network approach

    Lei Chen, Le Wu, Richang Hong, Kun Zhang, and Meng Wang. Revisiting graph based collab- orative filtering: A linear residual graph convolutional network approach. In The Thirty-Fourth AAAI Conference on Artificial Intelligence, AAAI 2020, The Thirty-Second Innovative Appli- cations of Artificial Intelligence Conference, IAAI 2020, The Tenth AAAI Symposium...

    work page 2020

  6. [6]

    Myers, and Jure Leskovec

    Eunjoon Cho, Seth A. Myers, and Jure Leskovec. Friendship and mobility: user movement in location-based social networks. In Chid Apt´ e, Joydeep Ghosh, and Padhraic Smyth, editors, Proceedings of the 17th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Diego, CA, USA, August 21-24, 2011 , pages 1082–1090. ACM, 2011. 20

    work page 2011

  7. [7]

    Unified denoising training for recommendation

    Haoyan Chua, Yingpeng Du, Zhu Sun, Ziyan Wang, Jie Zhang, and Yew-Soon Ong. Unified denoising training for recommendation. In Tommaso Di Noia, Pasquale Lops, Thorsten Joachims, Katrien Verbert, Pablo Castells, Zhenhua Dong, and Ben London, editors,Proceedings of the 18th ACM Conference on Recommender Systems, RecSys 2024, Bari, Italy, October 14-18, 2024 ...

    work page 2024

  8. [8]

    Ziwei Fan, Ke Xu, Zhang Dong, Hao Peng, Jiawei Zhang, and Philip S. Yu. Graph collaborative signals denoising and augmentation for recommendation. In Hsin-Hsi Chen, Wei-Jou (Edward) Duh, Hen-Hsen Huang, Makoto P. Kato, Josiane Mothe, and Barbara Poblete, editors, Pro- ceedings of the 46th International ACM SIGIR Conference on Research and Development in I...

    work page 2023

  9. [9]

    A survey of graph neural networks for recom- mender systems: Challenges, methods, and directions

    Chen Gao, Yu Zheng, Nian Li, Yinfeng Li, Yingrong Qin, Jinghua Piao, Yuhan Quan, Jianxin Chang, Depeng Jin, Xiangnan He, and Yong Li. A survey of graph neural networks for recom- mender systems: Challenges, methods, and directions. Trans. Recomm. Syst., 1(1):1–51, 2023

    work page 2023

  10. [10]

    Content augmented graph neural networks

    Fatemeh Gholamzadeh Nasrabadi, Amirhossein Kashani, Pegah Zahedi, and Mostafa Haghir Chehreghani. Content augmented graph neural networks. ACM Trans. Web , October 2024

    work page 2024

  11. [11]

    Heterophily-aware fair recommendation using graph convolu- tional networks,

    Nemat Gholinejad and Mostafa Haghir Chehreghani. Heterophily-aware fair recommendation using graph convolutional networks. CoRR, abs/2402.03365, 2024

    work page arXiv 2024

  12. [12]

    Disentangling popularity and quality: An edge classification approach for fair recommendation, 2025

    Nemat Gholinejad and Mostafa Haghir Chehreghani. Disentangling popularity and quality: An edge classification approach for fair recommendation, 2025

    work page 2025

  13. [13]

    Maxwell Harper and Joseph A

    F. Maxwell Harper and Joseph A. Konstan. The movielens datasets: History and context. ACM Trans. Interact. Intell. Syst. , 5(4):19:1–19:19, 2016

    work page 2016

  14. [14]

    Ruining He and Julian J. McAuley. Ups and downs: Modeling the visual evolution of fashion trends with one-class collaborative filtering. In Jacqueline Bourdeau, Jim Hendler, Roger Nkam- bou, Ian Horrocks, and Ben Y. Zhao, editors, Proceedings of the 25th International Conference on World Wide Web, WWW 2016, Montreal, Canada, April 11 - 15, 2016 , pages 50...

    work page 2016

  15. [15]

    Lightgcn: Simplifying and powering graph convolution network for recommendation

    Xiangnan He, Kuan Deng, Xiang Wang, Yan Li, Yong-Dong Zhang, and Meng Wang. Lightgcn: Simplifying and powering graph convolution network for recommendation. In Jimmy X. Huang, Yi Chang, Xueqi Cheng, Jaap Kamps, Vanessa Murdock, Ji-Rong Wen, and Yiqun Liu, editors, Proceedings of the 43rd International ACM SIGIR conference on research and development in In...

    work page 2020

  16. [16]

    Neural collaborative filtering

    Xiangnan He, Lizi Liao, Hanwang Zhang, Liqiang Nie, Xia Hu, and Tat-Seng Chua. Neural collaborative filtering. pages 173–182, 2017

    work page 2017

  17. [17]

    Session-based recommendations with recurrent neural networks

    Bal´ azs Hidasi, Alexandros Karatzoglou, Linas Baltrunas, and Domonkos Tikk. Session-based recommendations with recurrent neural networks. In Yoshua Bengio and Yann LeCun, editors, 4th International Conference on Learning Representations, ICLR 2016, San Juan, Puerto Rico, May 2-4, 2016, Conference Track Proceedings, 2016

    work page 2016

  18. [18]

    Wang-Cheng Kang and Julian J. McAuley. Self-attentive sequential recommendation. In IEEE International Conference on Data Mining, ICDM 2018, Singapore, November 17-20, 2018 , pages 197–206. IEEE Computer Society, 2018

    work page 2018

  19. [19]

    Kipf and Max Welling

    Thomas N. Kipf and Max Welling. Semi-supervised classification with graph convolutional net- works. In 5th International Conference on Learning Representations, ICLR 2017, Toulon, France, April 24-26, 2017, Conference Track Proceedings. OpenReview.net, 2017

    work page 2017

  20. [20]

    Bell, and Chris Volinsky

    Yehuda Koren, Robert M. Bell, and Chris Volinsky. Matrix factorization techniques for recom- mender systems. Computer, 42(8):30–37, 2009

    work page 2009

  21. [21]

    Jiacheng Li, Yujie Wang, and Julian J. McAuley. Time interval aware self-attention for sequential recommendation. In James Caverlee, Xia (Ben) Hu, Mounia Lalmas, and Wei Wang, editors, WSDM ’20: The Thirteenth ACM International Conference on Web Search and Data Mining, Houston, TX, USA, February 3-7, 2020 , pages 322–330. ACM, 2020

    work page 2020

  22. [22]

    TEA: A sequential recommendation framework via temporally evolving aggregations

    Zijian Li, Ruichu Cai, Fengzhu Wu, Sili Zhang, Hao Gu, Yuexing Hao, and Yuguang Yan. TEA: A sequential recommendation framework via temporally evolving aggregations. IEEE Trans. Neural Networks Learn. Syst. , 35(2):2628–2639, 2024

    work page 2024

  23. [23]

    Heterogeneous multidomain recommender system through adversarial learning

    Wenhui Liao, Qian Zhang, Bo Yuan, Guangquan Zhang, and Jie Lu. Heterogeneous multidomain recommender system through adversarial learning. IEEE Trans. Neural Networks Learn. Syst. , 34(11):8965–8977, 2023

    work page 2023

  24. [24]

    K-plet recurrent neural networks for sequential recommendation

    Xiang Lin, Shuzi Niu, Yiqiao Wang, and Yucheng Li. K-plet recurrent neural networks for sequential recommendation. In Kevyn Collins-Thompson, Qiaozhu Mei, Brian D. Davison, Yiqun Liu, and Emine Yilmaz, editors, The 41st International ACM SIGIR Conference on Research & Development in Information Retrieval, SIGIR 2018, Ann Arbor, MI, USA, July 08-12, 2018 ,...

    work page 2018

  25. [25]

    Selfgnn: Self-supervised graph neural networks for sequential recommendation

    Yuxi Liu, Lianghao Xia, and Chao Huang. Selfgnn: Self-supervised graph neural networks for sequential recommendation. In Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval , SIGIR ’24, page 1609–1618, New York, NY, USA, 2024. Association for Computing Machinery. 22

    work page 2024

  26. [26]

    Ultragcn: Ultra simplification of graph convolutional networks for recommendation

    Kelong Mao, Jieming Zhu, Xi Xiao, Biao Lu, Zhaowei Wang, and Xiuqiang He. Ultragcn: Ultra simplification of graph convolutional networks for recommendation. In Gianluca Demartini, Guido Zuccon, J. Shane Culpepper, Zi Huang, and Hanghang Tong, editors, CIKM ’21: The 30th ACM International Conference on Information and Knowledge Management, Virtual Event, Q...

    work page 2021

  27. [27]

    Factorizing personalized markov chains for next-basket recommendation

    Steffen Rendle, Christoph Freudenthaler, and Lars Schmidt-Thieme. Factorizing personalized markov chains for next-basket recommendation. In Michael Rappa, Paul Jones, Juliana Freire, and Soumen Chakrabarti, editors, Proceedings of the 19th International Conference on World Wide Web, WWW 2010, Raleigh, North Carolina, USA, April 26-30, 2010, pages 811–820....

    work page 2010

  28. [28]

    Unbi- ased recommender learning from missing-not-at-random implicit feedback

    Yuta Saito, Suguru Yaginuma, Yuta Nishino, Hayato Sakata, and Kazuhide Nakata. Unbi- ased recommender learning from missing-not-at-random implicit feedback. In James Caverlee, Xia (Ben) Hu, Mounia Lalmas, and Wei Wang, editors, WSDM ’20: The Thirteenth ACM Inter- national Conference on Web Search and Data Mining, Houston, TX, USA, February 3-7, 2020 , pag...

    work page 2020

  29. [29]

    Bert4rec: Sequential recommendation with bidirectional encoder representations from transformer

    Fei Sun, Jun Liu, Jian Wu, Changhua Pei, Xiao Lin, Wenwu Ou, and Peng Jiang. Bert4rec: Sequential recommendation with bidirectional encoder representations from transformer. In Wenwu Zhu, Dacheng Tao, Xueqi Cheng, Peng Cui, Elke A. Rundensteiner, David Carmel, Qi He, and Jeffrey Xu Yu, editors, Proceedings of the 28th ACM International Conference on Infor...

    work page 2019

  30. [30]

    Qingyun Sun, Jianxin Li, Hao Peng, Jia Wu, Xingcheng Fu, Cheng Ji, and Philip S. Yu. Graph structure learning with variational information bottleneck. In Thirty-Sixth AAAI Conference on Artificial Intelligence, AAAI 2022, Thirty-Fourth Conference on Innovative Applications of Ar- tificial Intelligence, IAAI 2022, The Twelveth Symposium on Educational Adva...

    work page 2022

  31. [31]

    Graph Convolutional Matrix Completion

    Rianne van den Berg, Thomas N. Kipf, and Max Welling. Graph convolutional matrix comple- tion. CoRR, abs/1706.02263, 2017

    work page internal anchor Pith review Pith/arXiv arXiv 2017

  32. [32]

    LLM4DSR: leveraing large language model for denoising sequential recommen- dation

    Bohao Wang, Feng Liu, Jiawei Chen, Yudi Wu, Xingyu Lou, Jun Wang, Yan Feng, Chun Chen, and Can Wang. LLM4DSR: leveraing large language model for denoising sequential recommen- dation. CoRR, abs/2408.08208, 2024

    work page arXiv 2024

  33. [33]

    Neural graph collabora- tive filtering

    Xiang Wang, Xiangnan He, Meng Wang, Fuli Feng, and Tat-Seng Chua. Neural graph collabora- tive filtering. In Benjamin Piwowarski, Max Chevalier, ´Eric Gaussier, Yoelle Maarek, Jian-Yun Nie, and Falk Scholer, editors, Proceedings of the 42nd International ACM SIGIR Conference 23 on Research and Development in Information Retrieval, SIGIR 2019, Paris, Franc...

    work page 2019

  34. [34]

    A survey on accuracy-oriented neural recommendation: From collaborative filtering to information-rich recommendation

    Le Wu, Xiangnan He, Xiang Wang, Kun Zhang, and Meng Wang. A survey on accuracy-oriented neural recommendation: From collaborative filtering to information-rich recommendation. IEEE Trans. Knowl. Data Eng. , 35(5):4425–4445, 2023

    work page 2023

  35. [35]

    Session-based recommendation with graph neural networks

    Shu Wu, Yuyuan Tang, Yanqiao Zhu, Liang Wang, Xing Xie, and Tieniu Tan. Session-based recommendation with graph neural networks. In The Thirty-Third AAAI Conference on Artifi- cial Intelligence, AAAI 2019, The Thirty-First Innovative Applications of Artificial Intelligence Conference, IAAI 2019, The Ninth AAAI Symposium on Educational Advances in Artifici...

    work page 2019

  36. [36]

    Enhancing robustness in implicit feedback recom- mender systems with subgraph contrastive learning

    Yi Yang, Shaopeng Guan, and Xiaoyang Wen. Enhancing robustness in implicit feedback recom- mender systems with subgraph contrastive learning. Inf. Process. Manag., 62(3):103962, 2025

    work page 2025

  37. [37]

    Re- drec: Relation and dynamic aware graph convolutional network for sequential recommendation

    Runfeng Yao, Weisheng Xu, Zhenyu Liu, Yang Wang, Zhen Li, Yuanyuan Qiao, and Jie Yang. Re- drec: Relation and dynamic aware graph convolutional network for sequential recommendation. In 8th IEEE International Conference on Network Intelligence and Digital Content, IC-NIDC 2023, Beijing, China, November 3-5, 2023 , pages 192–196. IEEE, 2023

    work page 2023

  38. [38]

    Denoised graph collaborative filtering via neighborhood similarity and dynamic thresholding

    Haibo Ye, Lijun Zhang, Yuan Yao, and Sheng-Jun Huang. Denoised graph collaborative filtering via neighborhood similarity and dynamic thresholding. IEEE Trans. Big Data , 10(6):683–693, 2024

    work page 2024

  39. [39]

    Dynamic graph neural networks for sequential recommendation

    Mengqi Zhang, Shu Wu, Xueli Yu, Qiang Liu, and Liang Wang. Dynamic graph neural networks for sequential recommendation. IEEE Trans. Knowl. Data Eng. , 35(5):4741–4753, 2023

    work page 2023

  40. [40]

    Dis- entangling long and short-term interests for recommendation

    Yu Zheng, Chen Gao, Jianxin Chang, Yanan Niu, Yang Song, Depeng Jin, and Yong Li. Dis- entangling long and short-term interests for recommendation. In Fr´ ed´ erique Laforest, Rapha¨ el Troncy, Elena Simperl, Deepak Agarwal, Aristides Gionis, Ivan Herman, and Lionel M´ edini, editors, WWW ’22: The ACM Web Conference 2022, Virtual Event, Lyon, France, Apri...

    work page 2022

  41. [41]

    Incorporating price into recom- mendation with graph convolutional networks

    Yu Zheng, Chen Gao, Xiangnan He, Depeng Jin, and Yong Li. Incorporating price into recom- mendation with graph convolutional networks. IEEE Trans. Knowl. Data Eng., 35(2):1609–1623, 2023

    work page 2023

  42. [42]

    Graph neural networks: A review of methods and applica- tions

    Jie Zhou, Ganqu Cui, Shengding Hu, Zhengyan Zhang, Cheng Yang, Zhiyuan Liu, Lifeng Wang, Changcheng Li, and Maosong Sun. Graph neural networks: A review of methods and applica- tions. AI Open, 1:57–81, 2020. 24

    work page 2020

  43. [43]

    Centrality- based and similarity-based neighborhood extension in graph neural networks

    Mohammadjavad Zohrabi, Saeed Saravani, and Mostafa Haghir Chehreghani. Centrality- based and similarity-based neighborhood extension in graph neural networks. J. Supercomput., 80(16):24638–24663, 2024. 25

    work page 2024