Pith. sign in

REVIEW 4 major objections 5 minor 1 cited by

DisCo: Graph-Based Disentangled Contrastive Learning for Cold-Start Cross-Domain Recommendation

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read The paper claims that cold-start cross-domain recommendation improves when user representations are split into intents and only the source intents that reproduce target-domain user similarity are transferred.

desk verdict A well-assembled architecture with honest ablations, but the cold-start evaluation tests the wrong population and the 'consistent SOTA' claim does not survive Table 1. read the letter →

arxiv 2412.15005 v3 pith:6ZBPOFXP submitted 2024-12-19 cs.IR cs.LG

Hourun Li , Yifan Wang , Zhiping Xiao , Jia Yang , Changling Zhou , Ming Zhang , Wei Ju This is my paper · ORCID
classification cs.IRcs.LG
keywords cross-domainrecommendationcold-startusersdisentangledrepresentationlearningcontrastivegraphneuralnetworksnegativetransferuserintentmodelingrandom-walksimilarity
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

Cold-start cross-domain recommendation usually transfers user embeddings from a source domain where data exist to a target domain where new users have none. The paper argues that this transfer is corrupted when users who look similar in the source domain have different tastes in the target domain, because the source embeddings carry domain-specific collaborative information that misleads the target. To fix this, DisCo represents each user as several distinct intents, builds affinity graphs over user embeddings, and uses multi-step random walks to define high-order user similarity in each domain. An intent-wise contrastive objective then transfers only those source intents that reproduce target-domain user similarity, leaving target-specific preferences intact. The paper reports that on four Amazon domain pairs this consistently improves over embedding-mapping and disentangled cross-domain baselines.

What carries the argument

The load-bearing object is the intent-wise contrastive learning pair: a multi-channel graph encoder and a momentum (EMA) target encoder produce user-intent embeddings $z_{i,k}$ and $\hat{z}_{i,k}$ for $K$ intents. User affinity graphs are built from these embeddings, with edge weights $R_{ij}=\exp(-\|\hat{z}_{i,k}-\hat{z}_{j,k}\|/\tau)$, and normalized into a random-walk transition matrix; multi-step random walks give the high-order similarity $T_k = \alpha I + (1-\alpha)\tilde{R}^d$. This similarity matrix serves as the pseudo-label for both intra-domain contrast and the inter-domain loss, where a decoder $g^{(s\to t)}$ projects source intents into the target space and a variational EM step (an ELBO with $q(k|u_j,u_i)$) aligns them with target similarities $T^t$. An orthogonality loss on the intent channels keeps the intents disentangled. Together these pieces let the model select which source information to transfer.

What would settle it

Take the overlapping users used to build pseudo-labels $T^t$ and hold out a random half. Check whether the target-domain similarities $T^t$ predicted from momentum embeddings of the other half match the held-out users' actual target interactions, for example by testing whether users ranked as similar share held-out items. If the pseudo-label similarities are no better than random at predicting held-out target preferences, the cross-domain bridge $g^{(s\to t)}$ is learning from a miscalibrated signal and the cold-start gains should disappear.

Watch

Extended reading notes

Core claim

On its own terms, the paper's discovery is that negative transfer in cold-start cross-domain recommendation can be avoided by making the transfer intent-wise instead of embedding-wise. The model learns K user-intent channels per domain through a graph encoder, forms affinity graphs in the embedding space, and takes multi-step random walks on those graphs to get high-order user similarity. The cross-domain decoder $g^{(s\to t)}$ maps source intents to target intents, and an inter-domain contrastive loss, solved with a variational EM approximation, requires the transferred intents of overlapping users to reproduce the target domain's user similarity. At inference, a cold-start user's score for a target item is the intent-weighted sum of inner products between transferred user intents and target item intents. The experiments on four domain pairs are presented as evidence that this filtering avoids negative transfer and outperforms prior embedding-mapping and disentangled CDR models.

Load-bearing premise

The transfer rests on trusting that user similarities computed from overlapping users' target-domain embeddings are a reliable guide, and that those overlapping users represent cold-start users well enough for the learned decoder to transfer.

Editorial extensions

If this is right

  • Users with no target-domain history can receive recommendations that reflect source-side intents actually shared with target users, rather than the average behavior of their source neighbors.
  • The same affinity-graph pseudo-labeling can be reused for any two domains with overlapping users, without assuming the domains share the same preference geometry.
  • The number of intents $K$ becomes a controllable trade-off: the paper's parameter study suggests dense domains benefit from more intents, so intent granularity can be tuned per domain.
  • The ablation results imply that each module, including the cross-domain decoder, the learned intent prior, orthogonality, and random-walk similarity, contributes to the final gain.

Reading between the lines

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

  • Because the pseudo-labels $T^t$ are built from momentum embeddings of overlapping users, a natural robustness check is to weight the contrastive loss by label confidence or to re-estimate similarities after a few cold-start interactions.
  • The framework is written for two domains, but the cross-domain decoder and intent-wise contrastive loss are modular; multiple source domains could be combined by gating or averaging the decoders.
  • A testable prediction follows from the paper's own logic: on domain pairs where source and target intents overlap the least, DisCo should show the largest advantage over embedding-level transfer, and the learned intent prior $p(k|u)$ should concentrate on the transferable intents.
  • The variational EM approximation depends on mini-batch sampling of overlapping users; with very few overlapping users, the estimated posterior $q(k|u_j,u_i)$ may be noisy, so performance gains may shrink as overlap size drops.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper proposes DisCo, a graph-based disentangled contrastive learning framework for cold-start cross-domain recommendation. The method uses a multi-channel graph encoder to produce K intent-specific user/item embeddings, constructs affinity graphs in the embedding space, and applies multi-step random walks to obtain high-order user similarity targets. It then combines intra-domain contrastive learning (with an EMA target encoder and orthogonality regularization) with an inter-domain contrastive loss that aligns source-domain intent embeddings to target-domain similarities via a learned cross-domain decoder. Experiments are reported on four Amazon domain pairs with HR@10 and NDCG@10, together with ablations, parameter sensitivity studies, and a case study. The paper claims that DisCo consistently outperforms state-of-the-art baselines and mitigates negative transfer for cold-start users.

Significance. If the proposed method works as claimed, the contribution is valuable: intent-level disentanglement and similarity-guided contrastive alignment are plausible mechanisms for reducing negative transfer in cross-domain recommendation, and the paper includes a released codebase, ablations, and parameter analyses. However, the evidence is not yet sufficient to establish the central claims. The evaluation protocol tests only overlapping users whose target-domain interactions are removed, not the true cold-start (source-only) users defined in the problem statement; several directly relevant recent baselines are omitted; and the abstract overstates the empirical results, since Table 1 shows DisCo behind CDRIB and UniCDR on several dataset/metric combinations with overlapping confidence intervals. The core method is coherent and the paper is readable, but these load-bearing gaps require a major revision.

major comments (4)
  1. [Problem Definition / Experimental Setup] The evaluation does not test the population defined in the problem statement. The problem definition states that cold-start CDR targets non-overlapping source-only users u in Us, but the Experimental Setup says: 'we randomly select 20% of overlapping users (i.e., those observed in both source and target domains) and treat them as cold-start users by removing their target domain interactions during testing and validation.' All reported cold-start users are therefore members of Uo, not Us. The cross-domain decoder g(s->t) and the pseudo-label matrix T^t (Eqs. 4, 9, 14-15) are built entirely from overlapping users, so generalization from Uo to true cold-start users is untested. Because the abstract's central claim is about avoiding negative transfer for cold-start users, this is a load-bearing gap. I recommend either evaluating on held-out source-only users (Us) or explicitly reframing the claims as 'simulated cold-start' and providing evidence that Uo is representative of Us in terms of degree and embedding geometry.
  2. [Abstract / Table 1 / Performance Analysis] The abstract's claim that DisCo 'consistently outperforms existing state-of-the-art baselines' is not supported by the reported data. In Table 1, CDRIB has higher HR@10 on Sport (11.10±0.29 vs. 10.72±0.32) and higher NDCG@10 on Elec (12.01±0.22 vs. 11.54±0.23); UniCDR has higher NDCG@10 on Sport (6.19±0.17 vs. 5.81±0.26) and higher NDCG@10 on Cloth (7.31±0.26 vs. 6.92±0.32). Confidence intervals overlap in every such comparison, and no significance tests are reported. The body text already tempers this to 'remains competitive' for sport-cloth and phone-electronic, so the abstract and the performance-analysis narrative should be revised to match the data, or statistical tests should be added to support the superiority claim.
  3. [Eq. (17), User Intent Adaptation and Prediction] Equation (17) defines the matching score as \hat{y}_{uv} = sigmoid(r_{uv}) = 1/(1 + exp(r_{uv})). With the standard definition of sigmoid, this is 1/(1 + exp(r_{uv})) = sigmoid(-r_{uv}), not sigmoid(r_{uv}). Combined with the binary cross-entropy loss in Eq. (18), positive pairs would be optimized by driving r_{uv} toward -\infty and negative pairs by driving it toward +\infty, which inverts the intended preference ordering from Eq. (16), where larger inner products should indicate stronger preference. If this is a typo, it should be corrected to 1/(1 + exp(-r_{uv})); if the implementation actually uses Eq. (17) as written, the training objective is mis-specified. Given that this equation is the bridge between the learned representations and the recommendation loss, it is load-bearing and needs to be clarified.
  4. [Related Work / Experiments (Baselines)] Several directly relevant recent state-of-the-art cold-start CDR methods are discussed in the related work but are absent from the comparison in Table 1. In particular, AIM (Li et al. 2024a), CDRNP (Li et al. 2024b), DisenCDR (Cao et al. 2022a), and GDCCDR (Liu et al. 2024) are all cited but not included as baselines. Since the paper claims to outperform existing state-of-the-art baselines and specifically targets negative transfer, omitting these methods weakens the empirical claim. The authors should include these baselines (or justify their exclusion, e.g., by showing that they are not applicable to the chosen experimental protocol).
minor comments (5)
  1. [Eq. (10)-(14)] The notation in the inter-domain contrastive objective is hard to parse: Eq. (10) writes 'H(T^t, p(uj|ui)) = -T^t_{ij} log Ep(k|ui)p(uj|ui,k)', but the sum over j in the cross entropy is not shown, and the expectation is not parenthesized. Please clarify the full objective and the relationship between Eq. (9) and Eq. (10).
  2. [Eq. (7)] Equation (7) uses absolute-value bars around matrices (|Z_k^T Z_k - I|). Please specify the matrix norm (e.g., Frobenius norm) that is actually used, since '|.|' is ambiguous for matrices.
  3. [Implementation Details] The EMA update rule for the target encoder is mentioned but the momentum coefficient is never reported. Please state the exact update rule and the chosen momentum value, since the method's stability depends on it.
  4. [Throughout] There are several typos and minor wording issues: 'filtering our irrelevant source domain collaborative information' should be 'filtering out', 'retrain the user similarity information' should be 'retain', 'we loose the representation constraint' should be 'we relax', 'comfirm' should be 'confirm', and the x-axis label in Figure 3 reads 'Viedo' instead of 'Video'.
  5. [Case Study] The case study in Table 3 is anecdotal and reports only two users. It is useful for illustration, but the claim that DisCo 'mitigates negative transfer' would be better supported by a quantitative analysis, e.g., measuring the agreement between source-domain and target-domain neighborhoods before and after the transfer.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the contrastive targets are self-distillation signals, and the reported performance gains are measured on held-out target interactions rather than on the contrastive targets themselves.

full rationale

DisCo's claimed derivation chain is self-contained. The multi-channel graph encoder, affinity-graph random walks (Eqs. 3-4), and contrastive losses (Eqs. 5, 9) define a training objective; the pseudo-labels T_k and T^t are computed from the EMA target encoder's own embeddings. This is self-distillation, i.e., a self-referential training signal, not a circular reduction: the paper's headline quantitative claims are not about predicting T_k or T^t but about recommendation quality on held-out target interactions (Eq. 18, Table 1), and those interactions are not used to construct the contrastive targets. The cross-domain decoder g(s->t) is trained on overlapping users and evaluated on a held-out 20% of overlapping users whose target interactions are removed, which is a standard supervised split; the reported gains are therefore not forced by construction. No load-bearing self-citation or imported uniqueness theorem appears; citations such as Lu et al. (2024) for random walks and Cao et al. (2023) for the BCE loss are routine methodology references, and the orthogonality constraint (Eq. 7) is a standard proxy for disentanglement rather than an assumed conclusion. The one substantive concern is external validity: the paper defines cold-start users as U_s but evaluates on masked overlapping users U_o, which may not generalize to true source-only users. This is a correctness/validity limitation, not a circularity in the derivation.

Assumptions & free parameters 7 free parameters · 4 assumptions · 2 invented entities

The method introduces several hyperparameters and latent constructs. The central empirical gains could be sensitive to K, d, alpha, beta, lambda, gamma, and tau. The main domain assumption is that intent channels and the overlapping-user decoder transfer cleanly to cold-start users; this is the part most directly tied to the negative-transfer claim.

free parameters (7)
  • K (number of intents) = tuned in [1,6]
    Controls disentanglement granularity; tuned per dataset, see Figure 3.
  • d (random walk steps) = tuned in [1,5]
    Controls high-order similarity in Tk, see Figure 4.
  • alpha (self-similarity weight) = tuned in [0.1,0.9]
    Balances self and pair similarities in Tk = alpha I + (1-alpha) R^d.
  • beta (inter/intra contrast weight) = tuned in [0,0.5]
    Balances Linter against Lintra+Lorth in Equation 19.
  • lambda (contrast vs recommendation weight) = tuned in [0,0.5]
    Balances contrastive and recommendation losses in Equation 19.
  • gamma (orthogonality weight) = not reported
    Weight for Lorth in Equation 19; no tuning details given.
  • tau (temperature) = not reported
    Temperature in Equations 3 and 6; no value given.
assumptions (4)
  • standard math Variational EM approximation of the inter-domain objective yields a valid ELBO (Equations 12-14).
    The paper uses Bayes' theorem and an ELBO to justify the contrastive objective; this is standard variational inference but not formally derived or checked.
  • domain assumption User preferences can be decomposed into K independent intent channels, and K is the same across source and target domains.
    The multi-channel encoder and Equation 7 enforce orthogonality, but no evidence that these channels correspond to real user intents.
  • domain assumption Embedding-space distances via exp(-||z_i-z_j||/tau) capture meaningful user similarity.
    The affinity graph R and random walks in Equations 3-4 depend on this similarity being informative for both domains.
  • domain assumption The MLP decoder trained on overlapping users generalizes to cold-start users.
    The evaluation removes target interactions for cold-start users, so the transfer mapping must extrapolate beyond the training distribution.
invented entities (2)
  • K user-intent channels
    purpose: Represent hidden user intents separately and enable intent-wise contrastive transfer.
    These channels are internal latent variables with no external falsifiable handle.
  • Intent prototypes {c_k}
    purpose: Define the prior p(k|u_i) over intents in Equation 11.
    Learned during training; no independent measurement.

how reviews work

0 comments
Cite this review

Pith. "Pith review of DisCo: Graph-Based Disentangled Contrastive Learning for Cold-Start Cross-Domain Recommendation." pith.science (2026). https://pith.science/paper/6ZBPOFXP

@misc{pith2026241215005,
  author       = {Pith},
  title        = {Pith review of: DisCo: Graph-Based Disentangled Contrastive Learning for Cold-Start Cross-Domain Recommendation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6ZBPOFXP}},
  note         = {Machine review of arXiv:2412.15005}
}
read the original abstract

Recommender systems are widely used in various real-world applications, but they often encounter the persistent challenge of the user cold-start problem. Cross-domain recommendation (CDR), which leverages user interactions from one domain to improve prediction performance in another, has emerged as a promising solution. However, users with similar preferences in the source domain may exhibit different interests in the target domain. Therefore, directly transferring embeddings may introduce irrelevant source-domain collaborative information. In this paper, we propose a novel graph-based disentangled contrastive learning framework to capture fine-grained user intent and filter out irrelevant collaborative information, thereby avoiding negative transfer. Specifically, for each domain, we use a multi-channel graph encoder to capture diverse user intents. We then construct the affinity graph in the embedding space and perform multi-step random walks to capture high-order user similarity relationships. Treating one domain as the target, we propose a disentangled intent-wise contrastive learning approach, guided by user similarity, to refine the bridging of user intents across domains. Extensive experiments on four benchmark CDR datasets demonstrate that DisCo consistently outperforms existing state-of-the-art baselines, thereby validating the effectiveness of both DisCo and its components.

Figures

Figures reproduced from arXiv: 2412.15005 by the authors.

Figure 1
Figure 1. An illustration of user similarity distortion in CDR. [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Illustration of the proposed framework DisCo. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Performance comparison w.r.t. different numbers [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Performance comparison w.r.t. different values of [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Cold-Start Recommendation towards the Era of Large Language Models (LLMs): A Comprehensive Survey and Roadmap

    cs.IR 2025-01 conditional novelty 4.0 of 10

    A comprehensive survey and roadmap that groups cold-start recommendation methods into four knowledge scopes and defines nine cold-start problem types.

Reference graph

Works this paper leans on

49 extracted references · 44 canonical work pages · cited by 1 Pith paper

  1. [1]

    Bengio, Y.; Courville, A.; and Vincent, P. 2013. Representation learning: A review and new perspectives. IEEE Transactions on Pattern Analysis and Machine Intelligence, 35(8): 1798--1828

    work page 2013

  2. [2]

    Cao, J.; Li, S.; Yu, B.; Guo, X.; Liu, T.; and Wang, B. 2023. Towards universal cross-domain recommendation. In Proceedings of the International ACM Conference on Web Search & Data Mining, 78--86

    work page 2023

  3. [3]

    Cao, J.; Lin, X.; Cong, X.; Ya, J.; Liu, T.; and Wang, B. 2022 a . Disencdr: Learning disentangled representations for cross-domain recommendation. In Proceedings of the International ACM SIGIR Conference on Research & Development in Information Retrieval, 267--277

    work page 2022

  4. [4]

    Cao, J.; Sheng, J.; Cong, X.; Liu, T.; and Wang, B. 2022 b . Cross-domain recommendation to cold-start users via variational information bottleneck. In Proceedings of the IEEE Conference on Data Engineering, 2209--2223

    work page 2022

  5. [5]

    Chen, X.; Duan, Y.; Houthooft, R.; Schulman, J.; Sutskever, I.; and Abbeel, P. 2016. Infogan: Interpretable representation learning by information maximizing generative adversarial nets. In Proceedings of the Conference on Neural Information Processing Systems, 2172--2180

    work page 2016

  6. [6]

    R.; Shen, D.; Malon, C.; Zhang, Y.; Li, Y.; and Carin, L

    Cheng, P.; Min, M. R.; Shen, D.; Malon, C.; Zhang, Y.; Li, Y.; and Carin, L. 2020. Improving disentangled text representation learning with information-theoretic guidance. In Proceedings of the Annual Meeting of the Association for Computational Linguistics, 7530--7541

    work page 2020

  7. [7]

    Guan, R.; Pang, H.; Giunchiglia, F.; Liang, Y.; and Feng, X. 2022. Cross-domain meta-learner for cold-start recommendation. IEEE Transactions on Knowledge and Data Engineering, 35(8): 7829--7843

    work page 2022

  8. [8]

    Guo, X.; Li, S.; Guo, N.; Cao, J.; Liu, X.; Ma, Q.; Gan, R.; and Zhao, Y. 2023. Disentangled representations learning for multi-target cross-domain recommendation. ACM Transactions on Information Systems, 41(4): 1--27

    work page 2023

Show all 49 references
  1. [9]

    Higgins, I.; Matthey, L.; Pal, A.; Burgess, C.; Glorot, X.; Botvinick, M.; Mohamed, S.; and Lerchner, A. 2016. beta-vae: Learning basic visual concepts with a constrained variational framework. In Proceedings of the International Conference on Learning Representations

    2016

  2. [10]

    Hsieh, C.-K.; Yang, L.; Cui, Y.; Lin, T.-Y.; Belongie, S.; and Estrin, D. 2017. Collaborative metric learning. In Proceedings of the Web Conference, 193--201

    2017

  3. [11]

    Hu, G.; Zhang, Y.; and Yang, Q. 2018. Conet: Collaborative cross networks for cross-domain recommendation. In Proceedings of the International Conference on Information and Knowledge Management, 667--676

    2018

  4. [12]

    Ju, W.; Mao, Z.; Yi, S.; Qin, Y.; Gu, Y.; Xiao, Z.; Wang, Y.; Luo, X.; and Zhang, M. 2024 a . Hypergraph-enhanced Dual Semi-supervised Graph Classification. In Proceedings of the International Conference on Machine Learning

    2024

  5. [13]

    Ju, W.; Qin, Y.; Qiao, Z.; Luo, X.; Wang, Y.; Fu, Y.; and Zhang, M. 2022. Kernel-based substructure exploration for next POI recommendation. In Proceedings of the International Conference on Data Mining, 221--230

    2022

  6. [14]

    Ju, W.; Wang, Y.; Qin, Y.; Mao, Z.; Xiao, Z.; Luo, J.; Yang, J.; Gu, Y.; Wang, D.; Long, Q.; et al. 2024 b . Towards Graph Contrastive Learning: A Survey and Beyond. arXiv preprint arXiv:2405.11868

    2024 arXiv

  7. [15]

    Ju, W.; Yi, S.; Wang, Y.; Long, Q.; Luo, J.; Xiao, Z.; and Zhang, M. 2024 c . A survey of data-efficient graph learning. In Proceedings of the International Joint Conference on Artificial Intelligence, 8104--8113

    2024

  8. [16]

    Ju, W.; Yi, S.; Wang, Y.; Xiao, Z.; Mao, Z.; Li, H.; Gu, Y.; Qin, Y.; Yin, N.; Wang, S.; et al. 2024 d . A survey of graph neural networks in real world: Imbalance, noise, privacy and ood challenges. arXiv preprint arXiv:2403.04468

    2024

  9. [17]

    Kang, S.; Hwang, J.; Lee, D.; and Yu, H. 2019. Semi-supervised learning for cross-domain recommendation to cold-start users. In Proceedings of the International Conference on Information and Knowledge Management, 1563--1572

    2019

  10. [18]

    Li, B.; Yang, Q.; and Xue, X. 2009. Can movies and books collaborate? cross-domain collaborative filtering for sparsity reduction. In Proceedings of the International Joint Conference on Artificial Intelligence, 2052--2057

    2009

  11. [19]

    Li, H.; Ma, W.; Sun, P.; Li, J.; Yin, C.; He, Y.; Xu, G.; Zhang, M.; and Ma, S. 2024 a . Aiming at the Target: Filter Collaborative Information for Cross-Domain Recommendation. In Proceedings of the International ACM SIGIR Conference on Research & Development in Information Re...

    2024

  12. [20]

    Li, H.; Wang, X.; Zhang, Z.; Yuan, Z.; Li, H.; and Zhu, W. 2021. Disentangled contrastive learning on graphs. In Proceedings of the Conference on Neural Information Processing Systems, 21872--21884

    2021

  13. [21]

    Li, P.; and Tuzhilin, A. 2020. Ddtcdr: Deep dual transfer cross domain recommendation. In Proceedings of the International ACM Conference on Web Search & Data Mining, 331--339

    2020

  14. [22]

    Li, X.; Sheng, J.; Cao, J.; Zhang, W.; Li, Q.; and Liu, T. 2024 b . CDRNP: Cross-Domain Recommendation to Cold-Start Users via Neural Process. In Proceedings of the International ACM Conference on Web Search & Data Mining, 378--386

    2024

  15. [23]

    Liang, X.; Li, D.; and Madden, A. 2020. Attributed network embedding based on mutual information estimation. In Proceedings of the International Conference on Information and Knowledge Management, 835--844

    2020

  16. [24]

    Liu, J.; Sun, L.; Nie, W.; Jing, P.; and Su, Y. 2024. Graph Disentangled Contrastive Learning with Personalized Transfer for Cross-Domain Recommendation. In Proceedings of the AAAI Conference on Artificial Intelligence, 8769--8777

    2024

  17. [25]

    Liu, M.; Li, J.; Li, G.; and Pan, P. 2020. Cross domain recommendation via bi-directional transfer graph collaborative filtering networks. In Proceedings of the International Conference on Information and Knowledge Management, 885--894

    2020

  18. [26]

    Lu, Y.; Lin, Y.; Yang, M.; Peng, D.; Hu, P.; and Peng, X. 2024. Decoupled contrastive multi-view clustering with high-order random walks. In Proceedings of the AAAI Conference on Artificial Intelligence, 14193--14201

    2024

  19. [27]

    Ma, J.; Cui, P.; Kuang, K.; Wang, X.; and Zhu, W. 2019 a . Disentangled graph convolutional networks. In Proceedings of the International Conference on Machine Learning, 4212--4221

    2019

  20. [28]

    Ma, J.; Zhou, C.; Cui, P.; Yang, H.; and Zhu, W. 2019 b . Learning disentangled representations for recommendation. In Proceedings of the Conference on Neural Information Processing Systems, 5712--5723

    2019

  21. [29]

    Man, T.; Shen, H.; Jin, X.; and Cheng, X. 2017. Cross-domain recommendation: An embedding and mapping approach. In Proceedings of the International Joint Conference on Artificial Intelligence, 2464--2470

    2017

  22. [30]

    Qin, Y.; Wang, Y.; Sun, F.; Ju, W.; Hou, X.; Wang, Z.; Cheng, J.; Lei, J.; and Zhang, M. 2023. DisenPOI: Disentangling Sequential and Geographical Influence for Point-of-Interest Recommendation. In Proceedings of the International ACM Conference on Web Search & Data Mining, 508--516

    2023

  23. [31]

    Qu, X.; Wang, Y.; Li, Z.; and Gao, J. 2024. Graph-Enhanced Prompt Learning for Personalized Review Generation. Data Science and Engineering, 1--16

    2024

  24. [32]

    Rendle, S.; Freudenthaler, C.; Gantner, Z.; and Schmidt-Thieme, L. 2012. BPR: Bayesian personalized ranking from implicit feedback. arXiv preprint arXiv:1205.2618

    2012 arXiv

  25. [33]

    B.; and Lauw, H

    Salah, A.; Tran, T. B.; and Lauw, H. 2021. Towards source-aligned variational models for cross-domain recommendation. In Proceedings of the ACM Conference on Recommender Systems, 176--186

    2021

  26. [34]

    P.; and Gordon, G

    Singh, A. P.; and Gordon, G. J. 2008. Relational learning via collective matrix factorization. In Proceedings of the International ACM SIGKDD Conference on Knowledge Discovery & Data Mining, 650--658

    2008

  27. [35]

    Wang, X.; He, X.; Wang, M.; Feng, F.; and Chua, T.-S. 2019. Neural graph collaborative filtering. In Proceedings of the International ACM SIGIR Conference on Research & Development in Information Retrieval, 165--174

    2019

  28. [36]

    Wang, X.; Jin, H.; Zhang, A.; He, X.; Xu, T.; and Chua, T.-S. 2020 a . Disentangled graph collaborative filtering. In Proceedings of the International ACM SIGIR Conference on Research & Development in Information Retrieval, 1001--1010

    2020

  29. [37]

    Wang, Y.; Li, Y.; Li, S.; Song, W.; Fan, J.; Gao, S.; Ma, L.; Cheng, B.; Cai, X.; Wang, S.; et al. 2022 a . Deep graph mutual learning for cross-domain recommendation. In International Conference on Database Systems for Advanced Applications, 298--305

    2022

  30. [38]

    Wang, Y.; Luo, X.; Chen, C.; Hua, X.-S.; Zhang, M.; and Ju, W. 2024. DisenSemi: Semi-supervised Graph Classification via Disentangled Representation Learning. arXiv preprint arXiv:2407.14081

    2024 arXiv

  31. [39]

    Wang, Y.; Qin, Y.; Han, Y.; Yin, M.; Zhou, J.; Yang, H.; and Zhang, M. 2022 b . Ad-aug: Adversarial data augmentation for counterfactual recommendation. In Joint European Conference on Machine Learning and Knowledge Discovery in Databases, 474--490

    2022

  32. [40]

    Wang, Y.; Qin, Y.; Sun, F.; Zhang, B.; Hou, X.; Hu, K.; Cheng, J.; Lei, J.; and Zhang, M. 2022 c . DisenCTR: Dynamic graph-based disentangled representation for click-through rate prediction. In Proceedings of the International ACM SIGIR Conference on Research & Development in...

    2022

  33. [41]

    Wang, Y.; Shen, J.; Song, Y.; Wang, S.; and Zhang, M. 2022 d . HE-SNE: Heterogeneous event sequence-based streaming network embedding for dynamic behaviors. In International Joint Conference on Neural Networks, 1--8

    2022

  34. [42]

    Wang, Y.; Song, Y.; Li, S.; Cheng, C.; Ju, W.; Zhang, M.; and Wang, S. 2022 e . DisenCite: Graph-Based Disentangled Representation Learning for Context-Specific Citation Generation. In Proceedings of the AAAI Conference on Artificial Intelligence, 11449--11458

    2022

  35. [43]

    Wang, Y.; Tang, S.; Lei, Y.; Song, W.; Wang, S.; and Zhang, M. 2020 b . Disenhan: Disentangled heterogeneous graph attention network for recommendation. In Proceedings of the International Conference on Information and Knowledge Management, 1605--1614

    2020

  36. [44]

    Zhang, M.; Liu, S.; and Wang, Y. 2020. STR-SA: Session-based thread recommendation for online course forum with self-attention. In IEEE Global Engineering Education Conference, 374--381

    2020

  37. [45]

    Zhu, Q.; Zhu, Z.; Wang, Y.; and Yu, H. 2016. Fuzzy-information-based robustness of interconnected networks against attacks and failures. Physica A: Statistical Mechanics and its Applications, 458: 194--203

    2016

  38. [46]

    Zhu, Y.; Ge, K.; Zhuang, F.; Xie, R.; Xi, D.; Zhang, X.; Lin, L.; and He, Q. 2021. Transfer-meta framework for cross-domain recommendation to cold-start users. In Proceedings of the International ACM SIGIR Conference on Research & Development in Information Retrieval, 1813--1817

    2021

  39. [47]

    Zhu, Y.; Tang, Z.; Liu, Y.; Zhuang, F.; Xie, R.; Zhang, X.; Lin, L.; and He, Q. 2022. Personalized transfer of user preferences for cross-domain recommendation. In Proceedings of the International ACM Conference on Web Search & Data Mining, 1507--1515

    2022

  40. [48]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all...

  41. [49]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.