Pith. sign in

REVIEW 5 major objections 6 minor 46 references

Combining social relations and interaction data in Recommender System with Graph Convolution Collaborative Filtering

T0 review · 5 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read This paper claims that propagating user embeddings through interaction, Jaccard-similarity, and friendship graphs matches or beats nine baselines on recall, precision, and NDCG@20 in four datasets.

desk verdict The central user-correlation matrix C is undefined as written (union/intersection, not Jaccard), so the paper's core signal is unreproducible; the rest is a modest, interesting ablation study that would need code, data, and variance reporting before I'd trust the gains. read the letter →

arxiv 2506.02834 v1 pith:ERM2CWPP submitted 2025-06-03 cs.IR

Tin T. Tran , Vaclav Snasel , Loc Tan Nguyen This is my paper
classification cs.IR
keywords RecommendersystemsSocialCollaborativefilteringGraphconvolutionalnetworksJaccardsimilarityUsercorrelationImplicitfeedback
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

The paper proposes a graph-convolution recommender, GCCF, that learns user representations from three graphs at once: user–item interactions, a quantized Jaccard similarity graph between users based on overlapping purchase histories, and a social friendship graph. The authors claim this three-signal model reports the best or tied-best recall, precision, and NDCG@20 among nine baselines—MF, SocialMF, TrustSVD, GCMC, NGCF, WiGCN, LightGCN, SocialLGN, and SEPT—across all four datasets tested (Gowalla, Librarything, Ciao, Epinions). They also report that the correlation graph makes training converge in far fewer epochs than LightGCN-style baselines. If the paper is right, a simple sum of three normalized graph propagations, with no transformation matrices or activation functions, captures influence signals that heavier social-recommendation designs miss, and the extra user-correlation signal is cheap to compute and effective.

What carries the argument

The central mechanism is multi-graph linear embedding propagation. Every layer, the user embedding aggregates three symmetrically normalized neighborhood sums—from the interaction matrix A, the social matrix S, and the correlation matrix C—and the item embedding aggregates only from A (Equation 9); the three user signals are combined by weighted sum and, in matrix form (Equation 14), by simple addition. The new object is C: the paper computes the user-correlation signal from the Jaccard overlap of users' item sets and then applies a fixed quantization function f (Table 2) that maps overlap values into five levels, so that strong overlaps strongly influence propagation while weak overlaps are zeroed out. The final embedding is the mean over K layers, and the trainable parameters are only the initial embeddings, with the BPR loss trained by Adam.

What would settle it

Recompute the Table 4 numbers with the correlation matrix C removed from propagation (its normalized adjacency set to zero); if recall, precision, or NDCG@20 on any dataset stays at or above the variant that uses only interaction and social signals, then C contributes nothing and the central claim fails. A second check: the formula printed as equation (6), with WU the union size and WI the intersection size, yields values of one or larger rather than Jaccard similarities in [0,1], so the quantization in Table 2 is not directly reproducible from the text alone and requires the authors' implementation.

Watch

Extended reading notes

Core claim

The central claim is that user embeddings should be built by linearly propagating through three symmetrically normalized adjacency matrices at the same time: the user–item interaction matrix A, the social friendship matrix S, and a user-correlation matrix C derived from the Jaccard overlap of users' item sets, with overlap values quantized into five influence levels (0, 0.005, 0.05, 0.5, 1) according to Table 2. At each layer the three user-signal embeddings are simply summed and the item embedding propagates only through A (Equations 9 and 14), and the final user and item embeddings are the mean of all layer embeddings, scored by inner product and trained with Bayesian personalized ranking. With these design choices—no feature-transformation matrices and no nonlinear activations—the model reports the best or tied-best recall, precision, and NDCG@20 among MF, SocialMF, TrustSVD, GCMC, NGCF, WiGCN, LightGCN, SocialLGN, and SEPT on Gowalla, Librarything, Ciao, and Epinions. The authors interpret this as evidence that the correlation graph contributes influence signals beyond what high-order connectivity in the interaction graph and raw friendship edges already provide, and that quantizing overlaps prevents the model from overfitting to noisy small co-occurrences.

Load-bearing premise

The load-bearing premise is that the degree of overlap between two users' purchased or rated items carries influence information that is not already present in the purchase graph or the friendship graph, and that adding the three propagation signals by simple summation is enough to turn that influence into better recommendations.

Editorial extensions

If this is right

  • On Gowalla, Librarything, Ciao, and Epinions, the full model reports the best or tied-best recall, precision, and NDCG@20 among the nine baselines in Table 4; the recall gains over the best baseline are roughly 0.08 to 0.68 percentage points.
  • Adding the correlation matrix shortens training: on Gowalla the full model converges in 440 epochs versus 1,140 for LightGCN and 1,090 for SEPT, with similar reductions on the other datasets.
  • The ablations show the two extra signals play different roles: on Gowalla social friendship contributes the larger share of the improvement, while on Librarything the interaction-correlation signal matters more and social data alone slightly lowers NDCG; combining both always gives the best result.
  • Removing both feature-transformation matrices and nonlinear activations is again confirmed as the right simplification: SocialLGN variants that drop only one of the two perform worse, and the proposed model uses neither.

Reading between the lines

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

  • Because C is computed once from the interaction matrix with simple matrix operations, the observed convergence speedup suggests a preprocessing-only enhancement that could be bolted onto other LightGCN-style recommenders; the paper does not test that transfer.
  • The quantization into five fixed levels is one regularization choice; a continuous Jaccard weight or a learned mapping from overlap to influence would separate the benefit of the similarity signal itself from the benefit of threshold smoothing.
  • Since the social matrix is optional, the same architecture could accept other user–user relations, such as trust scores or co-purchase at category level, in place of friendship edges; the paper leaves that substitution untested.
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

5 major / 6 minor

Summary. This manuscript presents GCCF, a graph convolution collaborative filtering model that augments a LightGCN-style user-item propagation with two additional user-user signals: a quantized Jaccard overlap matrix C computed from purchase/rating histories and a social friendship matrix S. The authors propose a preprocessing pipeline (10-core filtering and user retention by Jaccard overlap), a propagation layer that averages normalized signals from A, C, and S, and a BPR loss. They compare against nine baselines on Gowalla, Librarything, Ciao, and Epinions and report that the full model achieves the highest recall, precision, and NDCG@20, while also converging faster than LightGCN-based baselines.

Significance. If the empirical claims are correct, the contribution is modest but useful: it shows that a lightweight, transformation-free GCN can incorporate social and behavioral user-user correlation signals, and it provides an ablation separating the two signals. The paper compares against a broad set of baselines and reports per-dataset results, which is a strength. However, the central new signal C is not well defined as written, the reported gains are small and without variance estimates, and the preprocessing makes the evaluation non-standard. Because of these issues, the significance of the claims cannot be assessed from the manuscript in its current form.

major comments (5)
  1. [III-A-2, Eqs. (4)-(6), Table 2] Equations (4) and (5) define WU = |Ii ∪ Ij| and WI = |Ii ∩ Ij|. Equation (6) then sets C = f(WU ⊙ WI^{-1}), whose element is |union| / |intersection|, which is ≥ 1 whenever the intersection is nonempty and undefined when it is empty. Table 2, however, partitions only values in [0,1] as Jaccard indices, and the sentence after (6) calls WI 'the matrix that represents the union.' The literal reading cannot produce the C used in the experiments; if the intended formula is WI ⊙ WU^{-1}, the paper must state this, handle the zero-intersection case, and re-run the experiments, because every reported improvement over LightGCN/WiGCN is attributed to this matrix.
  2. [III-A-2 and IV-A] The manuscript does not state whether the correlation matrix C is constructed from the full preprocessed interaction matrix or only from the 80% training split used for learning. If C is built from all interactions before the 20% latest-interaction test split, then test labels are visible in the user-user overlap signal during training, which would invalidate the evaluation. The authors should specify that C, S, and the propagation graph are computed from training data only, and recompute all tables if necessary.
  3. [IV-B, Table 4] The text in Section IV-B states that 'our model with interaction embedding (Our w/ interaction) still gives the highest results' among the non-social models MF, GCMC, NGCF, LightGCN, and WiGCN. Table 4 contradicts this on Epinions: for recall, LightGCN achieves 0.0150 versus 0.0139 for 'Our w/ interact', and for NDCG@20, 0.0085 versus 0.0073. This claim needs to be corrected or the table corrected.
  4. [III-A-2, Table 2] The quantization function f in Table 2 uses hand-selected bin boundaries and typical values (0.0, 0.005, 0.05, 0.5, 1.0) with no stated selection procedure and no sensitivity analysis. Because the reported improvements of the full model over LightGCN are small (e.g., +7.3% recall on Gowalla, +4.7% on Librarything; interaction-only gains of +2.1% and +0.3%), it is not possible to tell whether the gains come from the proposed propagation or from threshold choices tuned to the datasets. The authors should justify the thresholds via validation and report sensitivity or ablation results.
  5. [III-A, Algorithm 1 and Section IV-A] The preprocessing is not reproducible as specified: the text says 10-core drops users with fewer than 10 interactions, but Algorithm 1 filters out items with fewer than 10 interactions; the user-retention step keeps users with the highest Jaccard distance to the full item set using a 'selected_ratio' taken from the original dataset, which substantially changes the evaluation population. The authors should release the exact processed datasets and code, and clarify whether all baseline results are produced on the same processed data; otherwise the comparison in Table 4 cannot be verified.
minor comments (6)
  1. [III-A, Algorithm 1] The 10-core description is inconsistent: the text says users with fewer than 10 interactions are dropped, while Algorithm 1 filters out items with fewer than 10 interactions; these are different operations and the discrepancy should be resolved.
  2. [III-B-5, Eq. (14)] Equation (14) uses notation eS, eC, and eR that is not defined consistently with (15), and the superscripts on E^{(k-1)} versus E^{(k)} appear to be off by one relative to the propagation equations in (9).
  3. [IV-A, Eq. (16)] The definitions of false positive and false negative in (16) are incorrect; false negative is an actual interaction not retrieved, not 'predicted interactions not being exist on the test set.' The ranking metrics (recall@20, NDCG@20) should be defined precisely.
  4. [IV-C-1] The text mentions 'SocialGCN' in item 3, but this model is not among the baselines and is not defined; the intended reference may be SEPT.
  5. [V-A] The conclusion contains a placeholder citation '[?]' that should be removed or replaced.
  6. [General] There are numerous typos and inconsistencies, including 'analyst' for 'analyze' in the introduction, 'Epinion' in the Table 4 header, and 'N DGC' in equation (17).

Circularity Check

0 steps flagged · score 2.0 of 10

No circular derivation: held-out evaluation and external baselines make the central comparison self-contained; the only circularity-adjacent concern is a non-load-bearing self-citation, plus an internal inconsistency in the definition of C that is a reproducibility defect, not a circularity.

full rationale

The paper's claim is empirical: GCCF combines interaction matrix A, quantized user-correlation matrix C, and social matrix S, and reports higher recall/precision/NDCG@20 than listed baselines (Table 4). The evaluation protocol splits interactions by timestamp into 80% training and 20% test (Section IV-A), so metrics are computed against interactions not used in the BPR loss. The propagation equations (9) and (14) are standard LightGCN-style aggregation, and no parameter is fitted to the test labels. No uniqueness theorem or external mathematical result is invoked to force the architecture. The cited prior work [34] is by the same first author and is used to motivate adding a weighted user influence matrix, but the contribution of that matrix is tested directly via ablations (Tables 7 and 8), so the self-citation is not load-bearing. Two non-circular concerns should be weighed. First, Section III-A-2/Equation (6) is internally inconsistent: with WU = |Ii ∪ Ij| and WI = |Ii ∩ Ij|, the expression WU ⊙ WI^{-1} equals |Ii ∪ Ij| / |Ii ∩ Ij|, which is ≥ 1 and undefined for disjoint sets, while Table 2 bins values in [0,1). The text at (6) also reverses the roles by calling WI the union matrix. Thus the actual C used in experiments is not specified as written, which is a reproducibility defect but not a circular reduction. Second, the 10-core and user-selection preprocessing is dataset-dependent, but it is applied uniformly to all models and does not by construction determine the ranking in Table 4. Overall, the derivation is self-contained; the score of 2 reflects only the minor non-load-bearing self-citation and the flagged specification defect, not a by-construction equivalence.

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

The central claim rests on empirical setup choices rather than theoretical axioms. The hand-picked quantization bins, the 10-core threshold, and the ratio-based user selection are free parameters not justified by data or theory; the fusion assumption (sum of three normalized propagations) is ad hoc; and the undirected treatment of social edges is a domain assumption. No invented entities are introduced.

free parameters (4)
  • Jaccard quantization thresholds = 0.0, 0.005, 0.05, 0.5, 1.0
    Hand-chosen cutoff values in Table 2 convert raw Jaccard similarity into five discrete influence levels; no sensitivity analysis or data-driven fitting is provided, yet the value of C depends entirely on these.
  • 10-core threshold = 10
    Users and items with fewer than 10 interactions are removed; standard in the literature but still a free choice affecting dataset density and all results.
  • selected_ratio (item/user ratio for user retention) = datasets dependent: 1.44, 1.60, 18.78, 11.95
    Algorithm 1 keeps q = p / selected_ratio users, where selected_ratio is set to the original item/user ratio per dataset; this controls how many users are kept and is not independently justified.
  • embedding size d and layer count K = d=64, K unspecified
    Embedding dimension is set to 64; number of propagation layers is taken from prior work without stating the value used in the experiments.
assumptions (4)
  • domain assumption High-order connectivity in the user-item graph encodes collaborative filtering signals usable by linear propagation
    Adopted from LightGCN and NGCF (references [7], [24]) and used as the foundation of the propagation rules in equation (9).
  • domain assumption Social friendship edges S are undirected and binary even when the source data (e.g., Epinions trust, Ciao friend lists) may be directed or weighted
    Section III-A-3 states 'we treat the relationship of any couple of users as equal'; this ignores directionality and trust levels.
  • ad hoc to paper The user-correlation matrix C, derived from Jaccard item overlaps quantized by Table 2, captures user influence complementary to friendship S
    This is the central modeling bet of the paper; no evidence is provided that the hand-picked buckets are meaningful, and equation (6) as written does not produce Jaccard values in [0,1].
  • ad hoc to paper Keeping only the most active users (highest Jaccard overlap with the item catalog) preserves a representative evaluation population
    Algorithm 1 selects users with the largest interaction sets; this is a preprocessing choice that changes the composition of all datasets and is not validated against other selection strategies.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Combining social relations and interaction data in Recommender System with Graph Convolution Collaborative Filtering." pith.science (2026). https://pith.science/paper/ERM2CWPP

@misc{pith2026250602834,
  author       = {Pith},
  title        = {Pith review of: Combining social relations and interaction data in Recommender System with Graph Convolution Collaborative Filtering},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ERM2CWPP}},
  note         = {Machine review of arXiv:2506.02834}
}
read the original abstract

A recommender system is an important subject in the field of data mining, where the item rating information from users is exploited and processed to make suitable recommendations with all other users. The recommender system creates convenience for e-commerce users and stimulates the consumption of items that are suitable for users. In addition to e-commerce, a recommender system is also used to provide recommendations on books to read, movies to watch, courses to take or websites to visit. Similarity between users is an important impact for recommendation, which could be calculated from the data of past user ratings of the item by methods of collaborative filtering, matrix factorization or singular vector decomposition. In the development of graph data mining techniques, the relationships between users and items can be represented by matrices from which collaborative filtering could be done with the larger database, more accurate and faster in calculation. All these data can be represented graphically and mined by today's highly developed graph neural network models. On the other hand, users' social friendship data also influence consumption habits because recommendations from friends will be considered more carefully than information sources. However, combining a user's friend influence and the similarity between users whose similar shopping habits is challenging. Because the information is noisy and it affects each particular data set in different ways. In this study, we present the input data processing method to remove outliers which are single reviews or users with little interaction with the items; the next proposed model will combine the social relationship data and the similarity in the rating history of users to improve the accuracy and recall of the recommender system.

Figures

Figures reproduced from arXiv: 2506.02834 by the authors.

Figure 1
Figure 1. FIGURE 1 [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. FIGURE 2 [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. FIGURE 3 [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: FIGURE 4 [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: with Yelp data set. FIGURE 4. Compare variants of SocialLGN on data set Gowalla. FIGURE 5. Compare variants of SocialLGN on data set Yelp2018. When removing either feature transformation matrices W1, W2, W3 or non-linear activation functions σ(.), the model performance…
Figure 6
Figure 6. Figure 6: FIGURE 6 [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]
Figure 7
Figure 7. Figure 7: FIGURE 7 [PITH_FULL_IMAGE:figures/full_fig_p010_7.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

46 extracted references · 35 canonical work pages

  1. [1]

    & Zhang, C

    Zhang, J., Lin, Z., Xiao, B. & Zhang, C. An optimized item-based col- laborative filtering recommendation algorithm. 2009 IEEE International Conference On Network Infrastructure And Digital Content. pp. 414-418 (2009)

    work page 2009

  2. [2]

    & Barman, A

    Tewari, A., Kumar, A. & Barman, A. Book recommendation system based on combine features of content based filtering, collaborative filtering and association rule mining. 2014 IEEE International Advance Computing Conference (IACC). pp. 500-503 (2014)

    work page 2014

  3. [3]

    & Alippi, C

    Grattarola, D. & Alippi, C. Graph Neural Networks in TensorFlow and Keras with Spektral [Application Notes]. IEEE Computational Intelligence Magazine. 16, 99-106 (2021)

    work page 2021

  4. [4]

    & King, I

    Ma, H., Yang, H., Lyu, M. & King, I. SoRec: Social Recommendation Using Probabilistic Matrix Factorization. Proceedings Of The 17th ACM Conference On Information And Knowledge Management. pp. 931-940 (2008), https://doi.org/10.1145/1458082.1458205

    work page arXiv 2008

  5. [5]

    & King, I

    Ma, H., Zhou, D., Liu, C., Lyu, M. & King, I. Recommender systems with social regularization. WSDM. pp. 287-296 (2011,1)

    work page 2011

  6. [6]

    & Portinale, L

    Dossena, M., Irwin, C. & Portinale, L. Graph-based Recommendation using Graph Neural Networks. 2022 21st IEEE International Conference On Machine Learning And Applications (ICMLA). pp. 1769-1774 (2022)

    work page 2022

  7. [7]

    Neural Graph Collaborative Filtering

    Wang, X., He, X., Wang, M., Feng, F. & Chua, T. Neural Graph Collaborative Filtering. CoRR. abs/1905.08108 (2019), http://arxiv.org/abs/1905.08108

    work page Pith review arXiv 2019

  8. [8]

    & Boullé, M

    Candillier, L., Meyer, F. & Boullé, M. Comparing State-of-the-Art Collab- orative Filtering Systems. MLDM. (2007)

    work page 2007

Show all 46 references
  1. [9]

    & Salehi, M

    Jalili, M., Ahmadian, S., Izadi, M., Moradi, P. & Salehi, M. Evaluating Collaborative Filtering Recommender Algorithms: A Survey. IEEE Ac- cess. 6 pp. 74003-74024 (2018)

    2018

  2. [10]

    A Framework for Collaborative, Content-Based and Demo- graphic Filtering

    Pazzani, M. A Framework for Collaborative, Content-Based and Demo- graphic Filtering. ARTIFICIAL INTELLIGENCE REVIEW. 13 pp. 393- 408 (1999)

    1999

  3. [11]

    & Tuzhilin, A

    Adomavicius, G., Sankaranarayanan, R., Sen, S. & Tuzhilin, A. Incorpo- rating contextual information in recommender systems using a multidi- mensional approach. ACM Transactions On Information Systems (TOIS). 23, 103-145 (2005)

    2005

  4. [12]

    & Moradi, P

    Bojnordi, E. & Moradi, P. A novel collaborative filtering model based on combination of correlation method with matrix completion technique. The 16th CSI International Symposium On Artificial Intelligence And Signal Processing (AISP 2012). pp. 191-194 (2012)

    2012

  5. [13]

    & Terry, D

    Goldberg, D., Nichols, D., Oki, B. & Terry, D. Using Collaborative Filtering to Weave an Information Tapestry. Commun. ACM. 35, 61-70 (1992,12), https://doi.org/10.1145/138859.138867

    1992

  6. [14]

    & Kriegel, H

    Yu, K., Xu, X., Tao, J., Ester, M. & Kriegel, H. Instance Selection Techniques for Memory-Based Collaborative Filtering. Proceedings Of The 2002 SIAM International Conference On Data Mining (SDM). pp. 59-74, https://epubs.siam.org/doi/abs/10.1137/1.9781611972726.4

    2002 doi

  7. [15]

    & V olinsky, C

    Koren, Y ., Bell, R. & V olinsky, C. Matrix Factorization Techniques for Recommender Systems. Computer. 42, 30-37 (2009)

    2009

  8. [16]

    & Welling, M

    Berg, R., Kipf, T. & Welling, M. Graph Convolutional Matrix Completion. (2017)

    2017

  9. [17]

    Guo, Q., Zhuang, F., Qin, C., Zhu, H., Xie, X., Xiong, H. & He, Q. A Survey on Knowledge Graph-Based Recommender Systems. IEEE Transactions On Knowledge; Data Engineering. 34, 3549-3568 (2022,8) VOLUME 4, 2016 11 Tin T. Tranet al.: Combining social relations and interaction da...

    2022

  10. [18]

    & Khan, L

    Gao, Y ., Li, Y ., Lin, Y ., Gao, H. & Khan, L. Deep Learning on Knowledge Graph for Recommender System: A Survey. (2020)

    2020

  11. [19]

    & Cui, B

    Wu, S., Sun, F., Zhang, W. & Cui, B. Graph Neural Networks in Recom- mender Systems: A Survey. (2021)

    2021

  12. [20]

    Wu, Z., Pan, S., Chen, F., Long, G., Zhang, C. & Yu, P. A Comprehensive Survey on Graph Neural Networks. IEEE Transac- tions On Neural Networks And Learning Systems. 32, 4-24 (2021,1), http://dx.doi.org/10.1109/TNNLS.2020.2978386

    2021

  13. [21]

    Wang, X., Ji, H., Shi, C., Wang, B., Cui, P., Yu, P. & Ye, Y . Het- erogeneous Graph Attention Network. CoRR. abs/1903.07293 (2019), http://arxiv.org/abs/1903.07293

    2019 arXiv

  14. [22]

    & Leskovec, J

    Ying, R., He, R., Chen, K., Eksombatchai, P., Hamilton, W. & Leskovec, J. Graph Convolutional Neural Networks for Web-Scale Recommender Systems. Proceedings Of The 24th ACM SIGKDD International Con- ference On Knowledge Discovery &; Data Mining. pp. 974-983 (2018), https://doi...

    2018

  15. [23]

    & Leskovec, J

    Hamilton, W., Ying, R. & Leskovec, J. Inductive Representa- tion Learning on Large Graphs. CoRR. abs/1706.02216 (2017), http://arxiv.org/abs/1706.02216

    2017 arXiv

  16. [24]

    & Wang, M

    He, X., Deng, K., Wang, X., Li, Y ., Zhang, Y . & Wang, M. LightGCN: Simplifying and Powering Graph Convolution Network for Recommenda- tion. (2020)

    2020

  17. [25]

    & Yorke-Smith, N

    Guo, G., Zhang, J. & Yorke-Smith, N. A Novel Recommendation Model Regularized with User Trust and Item Ratings. IEEE Transactions On Knowledge And Data Engineering. 28, 1607-1620 (2016)

    2016

  18. [26]

    & Yang, S

    Jiang, M., Cui, P., Wang, F., Zhu, W. & Yang, S. Scalable Recommendation with Social Contextual Information. IEEE Transactions On Knowledge And Data Engineering. 26, 2789-2802 (2014)

    2014

  19. [27]

    & Fowler, J

    Bond, R., Fariss, C., Jones, J., Kramer, A., Marlow, C., Set- tle, J. & Fowler, J. A 61-million-person experiment in social in- fluence and political mobilization. Nature. 489, 295-298 (2012,9,1), https://doi.org/10.1038/nature11421

    2012 doi

  20. [28]

    & Tang, J

    Qiu, J., Tang, J., Ma, H., Dong, Y ., Wang, K. & Tang, J. Deep- Inf : Modeling Influence Locality in Large Social Networks. (2018), https://api.semanticscholar.org/CorpusID:49239721

    2018

  21. [29]

    & Srihari, S

    Wang, J., Bagul, D. & Srihari, S. ContextMF : A Fast and Context-aware Embedding Learning Method for Recommendation Systems. (2018), https://api.semanticscholar.org/CorpusID:198186939

    2018

  22. [30]

    Factorization Meets the Neighborhood: A Multifaceted Col- laborative Filtering Model

    Koren, Y . Factorization Meets the Neighborhood: A Multifaceted Col- laborative Filtering Model. Proceedings Of The 14th ACM SIGKDD International Conference On Knowledge Discovery And Data Mining. pp. 426-434 (2008), https://doi.org/10.1145/1401890.1401944

    2008

  23. [31]

    & Yin, D

    Fan, W., Ma, Y ., Li, Q., He, Y ., Zhao, E., Tang, J. & Yin, D. Graph Neural Networks for Social Recommendation. (2019)

    2019

  24. [32]

    & Zeng, J

    Liao, J., Zhou, W., Luo, F., Wen, J., Gao, M., Li, X. & Zeng, J. SocialLGN: Light graph convolution network for social recommendation. Information Sciences. 589 pp. 595-607 (2022), https://www.sciencedirect.com/science/article/pii/S0020025522000019

    2022

  25. [33]

    & Hung, N

    Yu, J., Yin, H., Gao, M., Xia, X., Zhang, X. & Hung, N. Socially- Aware Self-Supervised Tri-Training for Recommendation. CoRR. abs/2106.03569 (2021), https://arxiv.org/abs/2106.03569

    2021 arXiv

  26. [34]

    & Snasel, V

    Tran, T. & Snasel, V . Improvement Graph Convolution Collaborative Filtering with Weighted Addition Input. Intelligent Information And Database Systems. pp. 635-647 (2022)

    2022

  27. [35]

    & McSherry, F

    Najork, M. & McSherry, F. Computing Information Retrieval Performance Measures Efficiently in the Presence of Tied Scores. 30th European Conference On IR Research (ECIR). (2008,4), https://www.microsoft.com/en-us/research/publication/computing- information-retrieval-performanc...

    2008

  28. [36]

    & Chua, T

    He, X., Liao, L., Zhang, H., Nie, L., Hu, X. & Chua, T. Neural Collabora- tive Filtering. (2017)

    2017

  29. [37]

    & Schmidt-Thieme, L

    Rendle, S., Freudenthaler, C., Gantner, Z. & Schmidt-Thieme, L. BPR: Bayesian Personalized Ranking from Implicit Feedback. (2012)

    2012

  30. [38]

    Kingma, D. & Ba, J. Adam: A Method for Stochastic Optimization. (2017)

    2017

  31. [39]

    & Leskovec, J

    Cho, E., Myers, S. & Leskovec, J. Friendship and Mobility: User Move- ment in Location-Based Social Networks. Proceedings Of The 17th ACM SIGKDD International Conference On Knowledge Discovery And Data Mining. pp. 1082-1090 (2011), https://doi.org/10.1145/2020408.2020579

    2011

  32. [40]

    & King, I

    Zhao, T., McAuley, J. & King, I. Improving Latent Factor Mod- els via Personalized Feature Projection for One Class Recommen- dation. Proceedings Of The 24th ACM International On Conference On Information And Knowledge Management. pp. 821-830 (2015), https://doi.org/10.1145/28...

    2015

  33. [41]

    & Liu, H

    Tang, J., Gao, H. & Liu, H. MTrust: Discerning Multi-Faceted Trust in a Connected World. Proceedings Of The Fifth ACM International Conference On Web Search And Data Mining. pp. 93-102 (2012), https://doi.org/10.1145/2124295.2124309

    2012

  34. [42]

    & Kim, S

    Reyhani Hamedani, M., Ali, I., Hong, J. & Kim, S. TrustRec: An ef- fective approach to exploit implicit trust and distrust relationships along with explicit ones for accurate recommendations. Computer Science And Information Systems. 18 pp. 93-114 (2021,1)

    2021

  35. [43]

    & Riedl, J

    Herlocker, J., Konstan, J., Terveen, L. & Riedl, J. Evaluating Collabo- rative Filtering Recommender Systems. ACM Trans. Inf. Syst.. 22, 5-53 (2004,1), https://doi.org/10.1145/963770.963772

    2004

  36. [44]

    & Riedl, J

    Sarwar, B., Karypis, G., Konstan, J. & Riedl, J. Application of Dimension- ality Reduction in Recommender System – A Case Study. (2000,8)

    2000

  37. [45]

    & Riedl, J

    Sarwar, B., Karypis, G., Konstan, J. & Riedl, J. Analysis of Rec- ommendation Algorithms for E-Commerce. Proceedings Of The 2nd ACM Conference On Electronic Commerce. pp. 158-167 (2000), https://doi.org/10.1145/352871.352887

    2000

  38. [46]

    & Ester, M

    Jamali, M. & Ester, M. A Matrix Factorization Technique with Trust Propagation for Recommendation in Social Networks. Proceedings Of The Fourth ACM Conference On Recommender Systems. pp. 135-142 (2010), https://doi.org/10.1145/1864708.1864736 TIN T. TRAN received the M.S. degr...

    2010

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.