Pith. sign in

REVIEW 3 major objections 5 minor 43 references

Far From Sight, Far From Mind: Inverse Distance Weighting for Graph Federated Recommendation

T0 review · 3 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A distance-based aggregation rule for user embeddings improves federated graph recommendation accuracy without significant privacy or speed cost.

desk verdict Readable, but the method never aggregates the same user's embeddings across clients; the headline claim is unsupported. read the letter →

arxiv 2507.01285 v1 pith:XXTOSGME submitted 2025-07-02 cs.LG cs.DCcs.IR

Aymen Rayane Khouas , Mohamed Reda Bouadjenek , Hakim Hacid , Sunil Aryal This is my paper · ORCID
classification cs.LGcs.DCcs.IR
keywords federatedlearninggraphneuralnetworksrecommendersystemsmodelaggregationuserembeddingsinversedistanceweightingpersonalizationrecommendation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

Dist-FedAvg is an aggregation rule for graph federated recommendation that weights each user's updated embedding by the inverse of its distance to other users' embeddings, so that users with similar learned representations contribute more to the global model. The paper claims this distance-aware averaging improves recommendation accuracy over standard aggregation baselines while slotting into existing federated graph frameworks with minimal changes. The method also keeps the anchor user—the original user around whom each client's local graph is expanded—in the aggregate via linear interpolation, so evolving updates cannot drown out the primary user. This matters because federated recommenders commonly aggregate user parameters with uniform averaging, ignoring the fact that user embeddings carry per-person taste information. If the claim holds, Dist-FedAvg offers a cheap, privacy-compatible upgrade path for personalized federated recommendation.

What carries the argument

The carrying mechanism is inverse-distance weighting with anchor interpolation. Distances between user embeddings are computed with Minkowski distance (Eq. 3), a tunable-$p$ generalization of Euclidean and Manhattan distance; weights are the inverse of those distances for users inside the same expanded client graph (Eq. 4); and a row-normalization step converts them into an averaging scheme (Eqs. 5-6). The anchor-user contribution is then blended back in by linear interpolation (Eq. 7), and the interpolation parameter $\alpha$ can be held fixed or decayed so that early, still-random embeddings do not drive the aggregation. The mechanism's job is to make the global user embedding a similarity-weighted summary of local updates rather than a uniform average.

What would settle it

Run the same federated graph recommender used in the experiments on the same five datasets with the weighting rule inverted, giving more weight to users whose embeddings are farther apart, and also with random weights drawn from the same distribution; if Dist-FedAvg fails to beat both, the reported gains are not due to user-similarity weighting.

Watch

Extended reading notes

Core claim

At the end of each federated round, the server receives updated user embeddings from selected clients. Dist-FedAvg computes a Minkowski distance matrix $D_{ij}$ between the previous round's global user embeddings, converts distances into averaging weights $W_{ij}=1/D_{ij}$ for users in the same expanded local graph (excluding self), normalizes by row sums, and forms a weighted embedding $e'_{u_i}$. It then interpolates $e'_{u_i}$ with the anchor user's own updated embedding through $e^{(r)}_{u_i}=\alpha e_{u_i,i}+(1-\alpha)e'_{u_i}$, with $\alpha$ fixed, decayed arithmetically, or decayed geometrically. The paper reports higher HR@10 than FedAvg, SimpleAvg, FedMedian, and FedAtt on all five datasets, and higher NDCG@10 on four of the five, while keeping training speed and privacy properties intact.

Load-bearing premise

The load-bearing premise is that the distance between two user embeddings at aggregation time tracks how similar those users really are; if embeddings do not place similar users close together, inverse-distance weighting injects noise rather than signal.

Editorial extensions

If this is right

  • Dist-FedAvg can replace the user-embedding aggregation step in existing graph federated recommendation pipelines, leaving local training and item aggregation untouched.
  • It reports higher HR@10 than FedAvg, SimpleAvg, FedMedian, and FedAtt on all five datasets, and higher NDCG@10 on four of the five.
  • The anchor interpolation keeps the central user's locally learned embedding influential, so the aggregation adapts to evolving user interactions without forgetting the primary user.
  • Distances can be computed on-device and only weights transmitted, so the scheme can be adapted to a gradient-based privacy-preserving variant.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Because the scheme only needs user embeddings and pairwise distances, the same weighted-averaging idea could be applied to non-graph federated recommenders such as matrix-factorization or neural collaborative filtering, which the paper notes but does not test.
  • A direct next experiment would pair Dist-FedAvg with clustering or random sampling of the distance matrix to cut the $O(n^2)$ pairwise cost of the weighting step at scale.
  • Applying the same user-derived weights to item-embedding aggregation is an untested consequence of the method's logic; the paper identifies it as future work.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The manuscript proposes Dist-FedAvg, a distance-based aggregation rule for user embeddings in a FedPerGNN-style graph federated recommendation framework. The rule assigns inverse-Minkowski-distance weights to users in a client's expanded subgraph, computes a normalized weighted average, and linearly interpolates the result with the anchor user's own updated embedding using a decaying alpha. The authors compare Dist-FedAvg against FedAvg, SimpleAvg, FedMedian, and FedAtt on five datasets, report NDCG@10 and HR@10 with 95% confidence intervals, and include ablations on client count, alpha decay, and item-aggregation pairing. They claim consistent outperformance and seamless plug-in compatibility.

Significance. If the method as described actually performed cross-client aggregation and consistently improved NDCG@10 and HR@10, it would be a useful, low-overhead contribution: the code is released, the protocol is simple, and the evaluation covers multiple datasets with confidence intervals. However, two load-bearing problems prevent this assessment. First, Equations (4)-(7) define an intra-client averaging operation that does not combine a user's embeddings across selected clients, so the headline claim that the method combines local updates from different clients is not supported by the algorithm. Second, the paper's own tables contradict the claim of consistent outperformance: on ML-1m and LastFM-2k, Dist-FedAvg is worse than a baseline on one of the two metrics. The ablations and CI reporting are good practice, but they do not repair the mismatch between the claimed federated aggregation and the specified equations.

major comments (3)
  1. [Section 4, Eqs. (4)-(7) and Algorithm 2] As written, Dist-FedAvg never performs cross-client aggregation for a given user. In Eq. (6), e'_{u_i} is a weighted sum over e_{u_j,i}, the embeddings of users in client c_i's expanded subgraph, with W_{ij} nonzero only when u_j is in U_i; Eq. (7) then blends this with the anchor copy e_{u_i,i}. The copies e_{u_i,j} that Algorithm 1 collects from other selected clients j are never accessed in the aggregation step. Consequently, the output embedding of user u_i in round r is determined exclusively by client c_i's local updates, and information from other clients' current-round updates enters only indirectly through the previous global model E(r-1)_u. This makes Dist-FedAvg an intra-client neighborhood smoother rather than a federated aggregation rule, so the Section 1 claim that it combines local updates from different clients, and the comparison with FedAvg/FedMedian/FedAtt, are not supported by the specified algorithm.
  2. [Section 5.2, Tables 3 and 4] The abstract and Section 6 state that Dist-FedAvg consistently outperforms all baselines, but the reported numbers do not support this claim. On ML-1m, Dist-FedAvg NDCG@10 = 0.1672 ± 0.0047 is below FedAvg's 0.1678 ± 0.0047; on LastFM-2k, Dist-FedAvg HR@10 = 0.461 ± 0.0225 is below FedMedian's 0.4804 ± 0.0228. Several other comparisons are within overlapping confidence intervals, so even the favorable means should be reported as mixed results or ties, not as consistent wins.
  3. [Section 4, Algorithm 2 lines 9-13] For users u_i not in C(r), the algorithm sets e(r)_{u_i} = e'_{u_i}. However, e'_{u_i} is defined by Eq. (6) in terms of e_{u_j,i}, the updated embeddings of client c_i's expanded users, and client c_i was not selected in round r, so no current-round embeddings from c_i exist in E(r)_u. The algorithm therefore does not specify what data is used to compute e'_{u_i} for unselected users; it needs an explicit fallback, such as the previous global embedding, to be executable under partial client participation.
minor comments (5)
  1. [Table 1 and Algorithm 1] The notation E(r)_u is overloaded: it is used both for the set of updated embeddings collected from selected clients and for the global aggregated embedding matrix. Please disambiguate these two objects, for example with E_tilde(r)_u for the collected per-client updates.
  2. [Figures 2, 3, 4, 5] Figures 2 through 5 contain corrupted or unreadable text in the submitted version (for example, axis labels rendered as sequences such as 'uni0000002f'); these figures need to be regenerated with proper font rendering before the paper can be evaluated on the ablation results.
  3. [Section 3.3, footnote 3] The footnote acknowledges that incorporating privacy-preserving components such as pseudo-item sampling and differential privacy may introduce noise; the main experiments are conducted in the simplified framework without those components, so the reported gains should be described as gains in that simplified setting rather than as guarantees for a fully privacy-preserving deployment.
  4. [Section 4, Eq. (7) and Section 5.3] The text says alpha 'starts at 1' during decay, but Eq. (8) uses a tunable alpha_0 and the ablation in Figure 3 includes a 'No Decay' condition; please clarify which alpha schedule and hyperparameter values were used to produce the main results in Tables 3 and 4.
  5. [Section 4, Eq. (5)] If a client's expanded subgraph contains no users other than the anchor, the normalization factor d_i in Eq. (5) is zero and Eq. (6) is undefined; the authors should specify how this degenerate case is handled.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: Dist-FedAvg is a heuristic aggregation rule whose empirical claims are benchmarked against external baselines and not derived from its own construction.

full rationale

Dist-FedAvg is presented as a heuristic aggregation rule, not as a derivation of a prediction from first principles. The weighting scheme (Eqs. 3-7) is defined directly as inverse-distance weighting of user embeddings, so the statement that the method assigns higher aggregation weights to users with similar embeddings restates the construction rather than deriving an independent result; this is a design choice, not circularity. The empirical comparison against FedAvg, SimpleAvg, FedMedian, and FedAtt is an external benchmark against standard methods and is not used to define the method. Hyperparameters such as the Minkowski exponent p, the decay parameters, and alpha are tuned per dataset with Bayesian optimization, but the reported gains are on held-out test splits and are not fitted quantities renamed as predictions. The paper contains no load-bearing self-citations: references to FedPerGNN, LightGCN, and aggregation surveys are independent prior work. A separate specification concern exists: Eq. 6 sums only over u_j in U_i, the users in the anchor client's expanded local subgraph, so the implementation as written may not actually combine the same user's embeddings across clients; however, that is a correctness or scope issue rather than a circular reduction of the kind considered here. Consequently, no circular step is exhibited.

Assumptions & free parameters 7 free parameters · 4 assumptions · 0 invented entities

The method relies on several tuned aggregation hyperparameters and on an unvalidated geometric assumption about user embeddings. No new physical or conceptual entities are introduced.

free parameters (7)
  • p = tuned via Bayesian optimization; fixed value not reported
    Exponent in the Minkowski distance (Eq. 3); controls how strongly distances are penalized. Tuned per dataset.
  • alpha_0 = not reported
    Initial interpolation weight in Eq. 8; starts the decay. Tuned per dataset.
  • alpha_T = not reported
    Lower threshold for alpha in Eq. 8; prevents the anchor from disappearing. Tuned per dataset.
  • gamma = not reported
    Decay rate in Eq. 8; controls how fast alpha decreases. Tuned per dataset.
  • z = not reported
    Number of rounds per decay step in Eq. 8. Tuned per dataset.
  • decay_type = selected per dataset
    Choice between arithmetic (Eq. 8) and geometric (Eq. 9) decay. Tuned per dataset.
  • warmup_rounds = not reported
    Optional number of warmup rounds where alpha = 1, described in Section 4. Tuned per dataset.
assumptions (4)
  • domain assumption Minkowski distance between user embeddings is a meaningful measure of user similarity for aggregation
    Equations 3 and 4 use inverse distance as the aggregation weight. The paper's own decay discussion concedes embeddings are uninformative in early rounds, so the assumption only holds after some training, with no direct evidence provided.
  • ad hoc to paper The anchor user's embedding should retain a tunable share of the aggregated result
    Equation 7 interpolates with the anchor embedding. This design choice is asserted heuristically, not derived from any optimization or requirement.
  • domain assumption Removing privacy mechanisms (pseudo-item sampling, differential privacy) does not change the relative ranking of aggregation methods
    Section 3.3 and footnote 3 state the simplification. The authors concede that adding such mechanisms may introduce noise that impacts embedding quality.
  • standard math LightGCN with BPR loss is an adequate local recommender for evaluating aggregation rules
    Section 5.1 uses LightGCN, a standard and widely used model, but only one backbone is tested.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Far From Sight, Far From Mind: Inverse Distance Weighting for Graph Federated Recommendation." pith.science (2026). https://pith.science/paper/XXTOSGME

@misc{pith2026250701285,
  author       = {Pith},
  title        = {Pith review of: Far From Sight, Far From Mind: Inverse Distance Weighting for Graph Federated Recommendation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XXTOSGME}},
  note         = {Machine review of arXiv:2507.01285}
}
read the original abstract

Graph federated recommendation systems offer a privacy-preserving alternative to traditional centralized recommendation architectures, which often raise concerns about data security. While federated learning enables personalized recommendations without exposing raw user data, existing aggregation methods overlook the unique properties of user embeddings in this setting. Indeed, traditional aggregation methods fail to account for their complexity and the critical role of user similarity in recommendation effectiveness. Moreover, evolving user interactions require adaptive aggregation while preserving the influence of high-relevance anchor users (the primary users before expansion in graph-based frameworks). To address these limitations, we introduce Dist-FedAvg, a novel distance-based aggregation method designed to enhance personalization and aggregation efficiency in graph federated learning. Our method assigns higher aggregation weights to users with similar embeddings, while ensuring that anchor users retain significant influence in local updates. Empirical evaluations on multiple datasets demonstrate that Dist-FedAvg consistently outperforms baseline aggregation techniques, improving recommendation accuracy while maintaining seamless integration into existing federated learning frameworks.

Figures

Figures reproduced from arXiv: 2507.01285 by the authors.

Figure 1
Figure 1. An illustration of the aggregation of embeddings for u2 using Dist-FedAvg. To ensure the weights lie between 0 and 1, we normalize them by a factor di , defined as: di = Xn j=0 Wij . (5) Then normalize our embeddings with the following equation: e ′ ui = 1 di . Xn j=0 euj,iWij . (6) Handling the Anchor User and Interpolation: In Equation 4, Wij = 0 when i = j, meaning the anchor user’s embedding is excluded from the… view at source ↗
Figure 2
Figure 2. Difference in performance for Dist-FedAvg between different number of selected clients [PITH_FULL_IMAGE:figures/full_fig_p012_2.png] view at source ↗
Figure 3
Figure 3. Comparison of geometric, arith￾metic, and no decay strategies for com￾puting α. 0.000 0.025 0.050 0.075 0.100 0.125 0.150 0.175 NDCG@10 Lastfm-2k ML-100k Amazon Digital Music ML-1m Dataset 0.1125 0.1725 0.0584 0.1672 0.1124 0.1725 0.0507 0.1657 0.1097 0.1741 0.0586 0.1656 0.0887 0.1791 0.0273 0.1648 0.0000 Performance of Item Aggregation Methods with distfedavg Item Aggregation Method FedAvg SimpleAvg FedMedian FedA… view at source ↗
Figures from the paper (1 more)
Figure 5
Figure 5. Figure 5: Heatmap analyzing the impact of using different user (y-axis) and item (x￾axis) aggregation methods. The heatmap intensity represents the NDCG@10 score when using a combination of user and item aggregation methods beddings might negatively affect performance. We invest…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 33 canonical work pages

  1. [1]

    ACM Transactions on Recommender Systems (Jun 2023).https: //doi.org/10.1145/3651168

    AliWaqar, Ammad-ud-dinMuhammad, ZhouXiangmin, ZhangYan, ShaoJie: Communication-Efficient Federated Neural Collaborative Filtering with Multi- Armed Bandits. ACM Transactions on Recommender Systems (Jun 2023).https: //doi.org/10.1145/3651168

    work page doi:10.1145/3651168 2023

  2. [2]

    1901.09888

    Ammad-ud-din, M., Ivannikova, E., Khan, S.A., Oyomno, W., Fu, Q., Tan, K.E., Flanagan, A.: Federated Collaborative Filtering for Privacy-Preserving Person- alized Recommendation System (Jan 2019).https://doi.org/10.48550/arXiv. 1901.09888

    arXiv doi:10.48550/arxiv.1901.09888 2019

  3. [3]

    IEEE Intelligent Systems 36(5), 11–20 (Sep 2021)

    Chai, D., Wang, L., Chen, K., Yang, Q.: Secure Federated Matrix Factorization. IEEE Intelligent Systems 36(5), 11–20 (Sep 2021). https://doi.org/10.1109/ MIS.2020.3014880

    work page arXiv 2021

  4. [4]

    Applied Soft Computing 111, 107700 (Nov 2021)

    Du, Y., Zhou, D., Xie, Y., Shi, J., Gong, M.: Federated matrix factorization for privacy-preserving recommender systems. Applied Soft Computing 111, 107700 (Nov 2021). https://doi.org/10.1016/j.asoc.2021.107700

    arXiv 2021

  5. [5]

    In: Machine Learning and Knowledge Discovery in Databases

    Flanagan, A., Oyomno, W., Grigorievskiy, A., Tan, K.E., Khan, S.A., Ammad- Ud-Din, M.: Federated Multi-view Matrix Factorization for Personalized Recom- mendations. In: Machine Learning and Knowledge Discovery in Databases. pp. 324–347. Springer International Publishing, Cham (2021).https://doi.org/10. 1007/978-3-030-67661-2_20

    work page 2021

  6. [6]

    ACM Trans

    Guo, G., Zhang, J., Yorke-Smith, N.: A Novel Evidence-Based Bayesian Similarity Measure for Recommender Systems. ACM Trans. Web10(2), 8:1–8:30 (May 2016). https://doi.org/10.1145/2856037

    work page doi:10.1145/2856037 2016

  7. [7]

    ACM Trans

    Harper, F.M., Konstan, J.A.: The MovieLens Datasets: History and Context. ACM Trans. Interact. Intell. Syst.5(4), 19:1–19:19 (Dec 2015). https://doi.org/10. 1145/2827872 Inverse Distance Weighting for Graph Federated Recommendation 15

    work page 2015

  8. [8]

    In: SIGIR

    He, X., Deng, K., Wang, X., Li, Y., Zhang, Y., Wang, M.: LightGCN: Simplifying and Powering Graph Convolution Network for Recommendation. In: SIGIR. pp. 639–648. ACM, New York, NY, USA (Jul 2020)

    work page 2020

Show all 43 references
  1. [9]

    He, X., Liao, L., Zhang, H., Nie, L., Hu, X., Chua, T.S.: Neural Collaborative Filtering. In: WWW. pp. 173–182. WWW ’17, Republic and Canton of Geneva, CHE (Apr 2017).https://doi.org/10.1145/3038912.3052569

    2017

  2. [10]

    48550/arXiv.2403.03952

    Hou, Y., Li, J., He, Z., Yan, A., Chen, X., McAuley, J.: Bridging Language and Items for Retrieval and Recommendation (Mar 2024).https://doi.org/10. 48550/arXiv.2403.03952

    2024 doi arXiv

  3. [11]

    In: CIKM

    Huang, P.S., He, X., Gao, J., Deng, L., Acero, A., Heck, L.: Learning deep struc- tured semantic models for web search using clickthrough data. In: CIKM. pp. 2333–

  4. [12]

    Egyptian Informatics Journal16(3), 261–273 (Nov 2015)

    Isinkaye, F.O., Folajimi, Y.O., Ojokoh, B.A.: Recommendation systems: Principles, methods and evaluation. Egyptian Informatics Journal16(3), 261–273 (Nov 2015). https://doi.org/10.1016/j.eij.2015.06.005

    2015 doi

  5. [13]

    In: IJCNN

    Ji, S., Pan, S., Long, G., Li, X., Jiang, J., Huang, Z.: Learning Private Neural Language Modeling with Attentive Aggregation. In: IJCNN. pp. 1–8 (Jul 2019). https://doi.org/10.1109/IJCNN.2019.8852464

    2019

  6. [14]

    Computer42(8), 30–37 (Aug 2009).https://doi.org/10.1109/ MC.2009.263

    Koren, Y., Bell, R., Volinsky, C.: Matrix Factorization Techniques for Recom- mender Systems. Computer42(8), 30–37 (Aug 2009).https://doi.org/10.1109/ MC.2009.263

    2009

  7. [15]

    Liang, D., Krishnan, R.G., Hoffman, M.D., Jebara, T.: Variational Autoencoders for Collaborative Filtering. In: WWW. pp. 689–698. WWW ’18, International World Wide Web Conferences Steering Committee, Republic and Canton of Geneva, CHE (Apr 2018).https://doi.org/10.1145/3178876.3186150

    2018

  8. [16]

    IEEE Internet Computing 28(3), 17–24 (May 2024)

    Liu, Q., Yang, L., Liu, Y., Deng, J., Wu, G.: Privacy-Preserving Recommendation Based on a Shuffled Federated Graph Neural Network. IEEE Internet Computing 28(3), 17–24 (May 2024)

    2024

  9. [17]

    ACM Transactions on Intelligent Systems and Technology 13(4) (2022)

    Liu, Z., Yang, L., Fan, Z., Peng, H., Yu, P.: Federated Social Recommendation with Graph Neural Network. ACM Transactions on Intelligent Systems and Technology 13(4) (2022). https://doi.org/10.1145/3501815

    2022 doi

  10. [18]

    Decision Support Systems74, 12–32 (Jun 2015).https: //doi.org/10.1016/j.dss.2015.03.008

    Lu, J., Wu, D., Mao, M., Wang, W., Zhang, G.: Recommender system application developments: A survey. Decision Support Systems74, 12–32 (Jun 2015).https: //doi.org/10.1016/j.dss.2015.03.008

    2015 doi

  11. [19]

    In: AISTATS

    McMahan, B., Moore, E., Ramage, D., Hampson, S., y Arcas, B.A.: Communication-Efficient Learning of Deep Networks from Decentralized Data. In: AISTATS. pp. 1273–1282. PMLR (Apr 2017)

    2017

  12. [20]

    In: Proceedings of the 14th ACM Conference on Recommender Systems

    Meng, Z., McCreadie, R., Macdonald, C., Ounis, I.: Exploring Data Splitting Strategies for the Evaluation of Recommendation Models. In: Proceedings of the 14th ACM Conference on Recommender Systems. pp. 681–686. RecSys ’20, Asso- ciation for Computing Machinery, New York, NY, ...

    2020

  13. [21]

    In: SIGKDD

    Muhammad, K., Wang, Q., O’Reilly-Morgan, D., Tragos, E., Smyth, B., Hurley, N., Geraci, J., Lawlor, A.: FedFast: Going Beyond Average for Faster Training of Federated Recommender Systems. In: SIGKDD. pp. 1234–1242. KDD ’20, Associ- ation for Computing Machinery, New York, NY, ...

    2020

  14. [22]

    Khouas et al

    Ni, J., Li, J., McAuley, J.: Justifying Recommendations using Distantly-Labeled ReviewsandFine-GrainedAspects.In:EMNLP-IJCNLP.pp.188–197.Association for Computational Linguistics„ Hong Kong, China (Nov 2019) 16 A.R. Khouas et al

    2019

  15. [23]

    Knowledge-Based Systems 242, 108441 (Apr 2022)

    Perifanis, V., Efraimidis, P.S.: Federated Neural Collaborative Filtering. Knowledge-Based Systems 242, 108441 (Apr 2022). https://doi.org/10.1016/ j.knosys.2022.108441

    2022

  16. [24]

    Future Generation Computer Systems 150, 272–293 (Jan 2024)

    Qi, P., Chiaro, D., Guzzo, A., Ianni, M., Fortino, G., Piccialli, F.: Model aggrega- tion techniques in federated learning: A comprehensive survey. Future Generation Computer Systems 150, 272–293 (Jan 2024)

    2024

  17. [25]

    In: The Web Conference

    Qu, L., Tang, N., Zheng, R., Nguyen, Q.V.H., Huang, Z., Shi, Y., Yin, H.: Semi- decentralized Federated Ego Graph Learning for Recommendation. In: The Web Conference. pp. 339–348. WWW ’23, ACM, New York, NY, USA (Apr 2023)

    2023

  18. [26]

    In: The Web Conference

    Qu, L., Yuan, W., Zheng, R., Cui, L., Shi, Y., Yin, H.: Towards Personalized Privacy: User-Governed Data Contribution for Federated Recommendation. In: The Web Conference. pp. 3910–3918. WWW ’24, ACM, New York, NY, USA (May 2024)

    2024

  19. [27]

    Rendle, S., Freudenthaler, C., Gantner, Z., Schmidt-Thieme, L.: BPR: Bayesian personalized ranking from implicit feedback. In: UAI. pp. 452–461. AUAI Press, Arlington, Virginia, USA (Jun 2009)

    2009

  20. [28]

    Sedhain, S., Menon, A.K., Sanner, S., Xie, L.: AutoRec: Autoencoders Meet Col- laborative Filtering. In: WWW. pp. 111–112. WWW ’15 Companion, Association for Computing Machinery, New York, NY, USA (May 2015)

    2015

  21. [29]

    In: RecSys

    Shi, Y., Larson, M., Hanjalic, A.: Exploiting user similarity based on rated-item pools for improved user-based collaborative filtering. In: RecSys. pp. 125–132. RecSys ’09, ACM, New York, NY, USA (Oct 2009).https://doi.org/10.1145/ 1639714.1639736

    2009

  22. [30]

    IEEE Transactions on Neural Networks and Learning Systems pp

    Sun, Z., Xu, Y., Liu, Y., He, W., Kong, L., Wu, F., Jiang, Y., Cui, L.: A Survey on Federated Recommendation Systems. IEEE Transactions on Neural Networks and Learning Systems pp. 1–15 (2024).https://doi.org/10.1109/TNNLS.2024. 3354924

    2024 doi

  23. [31]

    Wang, X., He, X., Wang, M., Feng, F., Chua, T.S.: Neural Graph Collaborative Filtering.In:SIGIR.pp.165–174.SIGIR’19,AssociationforComputingMachinery, New York, NY, USA (Jul 2019)

    2019

  24. [32]

    In: SIGIR

    Wei, Y., Liu, W., Liu, F., Wang, X., Nie, L., Chua, T.S.: LightGT: A Light Graph Transformer for Multimedia Recommendation. In: SIGIR. pp. 1508–1517. SIGIR ’23, Association for Computing Machinery, New York, NY, USA (Jul 2023)

    2023

  25. [33]

    Nature Communications 13(1), 3091 (Jun 2022)

    Wu, C., Wu, F., Lyu, L., Qi, T., Huang, Y., Xie, X.: A federated graph neural net- work framework for privacy-preserving personalization. Nature Communications 13(1), 3091 (Jun 2022)

    2022

  26. [34]

    ACM Comput

    Wu, S., Sun, F., Zhang, W., Xie, X., Cui, B.: Graph Neural Networks in Recom- mender Systems: A Survey. ACM Comput. Surv.55(5), 97:1–97:37 (Dec 2022)

    2022

  27. [35]

    In: The Web Con- ference

    Yan, B., Cao, Y., Wang, H., Yang, W., Du, J., Shi, C.: Federated Heterogeneous Graph Neural Network for Privacy-preserving Recommendation. In: The Web Con- ference. pp. 3919–3929. WWW ’24, ACM, New York, NY, USA (May 2024)

    2024

  28. [36]

    In: Trustworthy Federated Learning

    Yang, L., Zhang, J., Chai, D., Wang, L., Guo, K., Chen, K., Yang, Q.: Practical and Secure Federated Recommendation with Personalized Masks. In: Trustworthy Federated Learning. Springer International Publishing (Jun 2022).https://doi. org/10.1007/978-3-031-28996-5_3

    2022 doi

  29. [37]

    In: ICML

    Yin, D., Chen, Y., Kannan, R., Bartlett, P.: Byzantine-Robust Distributed Learn- ing: Towards Optimal Statistical Rates. In: ICML. pp. 5650–5659. PMLR (Jul 2018)

    2018

  30. [38]

    Yin, H., Qu, L., Chen, T., Yuan, W., Zheng, R., Long, J., Xia, X., Shi, Y., Zhang, C.: On-Device Recommender Systems: A Comprehensive Survey (Jan 2024) Inverse Distance Weighting for Graph Federated Recommendation 17

    2024

  31. [39]

    https://doi.org/10.48550/arXiv.2008.07759

    Ying, S.: Shared MF: A privacy-preserving recommendation system (Aug 2020). https://doi.org/10.48550/arXiv.2008.07759

    2020 doi arXiv

  32. [40]

    In: SIGKDD

    Zhang, C., Long, G., Zhou, T., Zhang, Z., Yan, P., Yang, B.: GPFedRec: Graph- Guided Personalization for Federated Recommendation. In: SIGKDD. pp. 4131–

  33. [41]

    In: ICME

    Zhang, R., Chen, Y., Wu, C., Wang, F.: Cluster-driven GNN-based Federated Recommendation with Biased Message Dropout. In: ICME. pp. 594–599 (Jul 2023)

    2023

  34. [2338]

    https://doi.org/10.1145/2505515.2505665

    CIKM ’13, Association for Computing Machinery, New York, NY, USA (Oct 2013). https://doi.org/10.1145/2505515.2505665

    2013

  35. [4142]

    KDD ’24, ACM, New York, NY, USA (Aug 2024)

    2024

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.