Pith. sign in

REVIEW 4 major objections 6 minor 28 references

A Centralized-Distributed Transfer Model for Cross-Domain Recommendation Based on Multi-Source Heterogeneous Transfer Learning

T0 review · 4 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read CDTM claims that a centralized-distributed dual-embedding architecture lets a target domain borrow from many source domains at once, fixing feature and latent-space heterogeneity and lifting ad CTR by 5.1% online.

desk verdict A credible industrial multi-source CDR paper whose multi-source claim is supported by same-architecture comparisons, but whose central transfer-matrix mechanism has an internal consistency problem the authors never test. read the letter →

arxiv 2411.09286 v1 pith:2GDWV7HY submitted 2024-11-14 cs.LG

Ke Xu , Ziliang Wang , Wei Zheng , Yuhao Ma , Chenglin Wang , Nengxue Jiang , Cai Cao This is my paper · ORCID
classification cs.LG
keywords cross-domainrecommendationclick-throughratepredictionmulti-sourcetransferlearningheterogeneousdualembeddingcombinationattentionnegativeadvertisingsystem
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 CDTM, a centralized-distributed transfer model that lets a target advertising domain borrow knowledge from several source domains at once, rather than from a single source. It argues that the two main causes of negative transfer in cross-domain recommendation, different feature fields across domains and mismatched learned feature spaces, can be addressed by a dual embedding structure plus a learned transfer matrix and attention weights. The model is trained jointly across domains: each domain keeps its own domain-specific embedding, while all domains share a global shared embedding, and the target model adaptively mixes the two. Offline experiments on commercial advertising data report AUC gains over the single-domain baseline and over single-source transfer baselines on every tested target domain, and a two-week online A/B test reports 5.1% higher CTR and 6.6% higher eCPM than a DCN baseline.

What carries the argument

The load-bearing mechanism is the dual embedding structure combined with a linear transfer matrix and an attention gate. Each domain has a domain-specific embedding (DSE) table for its unique features, and all domains share a global shared embedding (GSE) table for transferable features; nontransferable features use only DSE, while transferable features get both. The combination layer merges the two embeddings for each transferable feature with $E = E_c \otimes A + T \otimes G_c \otimes (1-A)$, where the transfer matrix $T$ maps GSE into DSE's latent space to fix latent-space heterogeneity and the attention vector $A$, computed from $[E_c, E_c \otimes G_c, E_c \oplus G_c, G_c]$, weights which embedding matters more per feature. The auxiliary loss drives $T \otimes G_c$ toward $E_c$, and the total loss is a weighted sum over all domains of prediction loss plus this alignment term.

What would settle it

Run CDTM with the transfer matrix replaced by a per-field nonlinear mapping, or with the auxiliary alignment loss removed, and compare AUC on the same F1-F4 targets; if the gains over single-source transfer persist, the claimed latent-space-alignment mechanism is not what drives the results. Alternatively, inspect the learned embeddings directly: if $T \otimes G_c$ nearly equals $E_c$ for most features, the auxiliary loss has collapsed the two spaces and the attention weights carry the entire effect.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that a symmetric centralized-distributed architecture can make multi-source, heterogeneous cross-domain transfer work in CTR prediction. For each transferable feature, CDTM combines the domain-specific embedding $E_c$ with the global shared embedding $G_c$ through $E = E_c \otimes A + T \otimes G_c \otimes (1-A)$, where $T$ is a transfer matrix that maps $G_c$ into the latent space of $E_c$ and $A$ is a learned combination-attention vector. An auxiliary loss $\lambda \sum \|E_c - T \otimes G_c\|^2$ is added so the transfer matrix is optimized to bring the two embedding spaces into alignment. The paper reports that this structure beats single-source variants on all four target domains, avoids the negative transfer seen in several baselines, and extends to a four-domain jointly trained setup that improves all four domains at once.

Load-bearing premise

The model assumes a single learned linear element-wise transfer matrix can align the global shared embedding with each domain's own embedding, and that the alignment loss does not just collapse the shared embedding into the domain-specific one.

Editorial extensions

If this is right

  • Multi-source transfer consistently beats single-source transfer: CDTM outperforms both CDTM-H and CDTM-J on every target domain in Table IV.
  • Handling heterogeneity prevents negative transfer: several baselines lose AUC on at least one target domain, while CDTM improves on all of them in every configuration.
  • The symmetric design extends to multi-target scenarios: a four-domain jointly trained CDTM 4 improves all four domains simultaneously over the single-domain base model.
  • The method survives deployment: a two-week online A/B test reports 5.1% CTR lift and 6.6% eCPM lift over the DCN baseline.
  • Both components matter: removing combination attention still beats the base model, and adding it back improves results further in the ablation study.

Reading between the lines

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

  • A natural next step the paper does not take is to measure whether each source domain's contribution is positive online; the reported A/B test compares only against DCN, not against single-source variants, so per-source negative transfer in production remains untested.
  • Because the transfer matrix is shared element-wise across all transferable features, a per-field or per-feature transfer map would be a direct extension and could capture cases where the relation between global and domain spaces differs across features.
  • The gains reported on commercial data leave open how the method behaves on public multi-domain benchmarks; a testable extension is to apply CDTM to a public dataset with more than two source domains and report confidence intervals across seeds.
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 / 6 minor

Summary. The paper proposes CDTM, a centralized-distributed transfer model for multi-source cross-domain CTR prediction. The architecture gives each domain a domain-specific embedding (DSE) and all domains a jointly trained global shared embedding (GSE). Transferable features are represented by both embeddings, and the combination layer maps GSE through a transfer matrix T, then adaptively combines it with DSE using an attention vector A (Eq. (1)). The transfer matrix is intended to align the heterogeneous latent spaces of DSE and GSE by minimizing the Euclidean distance in Eq. (2), and an auxiliary loss version of this objective is added to the total training loss in Eqs. (7)-(8). Experiments on a proprietary NetEase Cloud Music advertising dataset compare CDTM with DCN, CoNet, SCoNet, DDTCDR, DTCDR, and GA-DTCDR under single-source and multi-source settings, include a four-domain extension, and report an online A/B test with 5.1% CTR and 6.6% eCPM improvements over DCN.

Significance. If the central claims were fully established, the paper would make a useful empirical contribution to multi-source heterogeneous cross-domain recommendation. The strongest evidence is Table IV, where CDTM is compared against CDTM-H and CDTM-J under the same architecture: this directly supports the multi-source-over-single-source claim. The paper also usefully identifies feature-dimensional and latent-space heterogeneities that many earlier CDR methods ignore, and the online deployment is an encouraging practical signal. However, the central transfer-matrix mechanism is not empirically isolated. There is no same-architecture source-free baseline, no ablation that removes the transfer matrix or the auxiliary loss, no report of the achieved alignment distance, no diagnostic of the learned attention weights, and no error bars or significance tests on the offline AUC comparisons. Because the reported gaps are often very small (for example, 0.0021 between CDTM and GA-DTCDR-H on F2 in Table II), the evidence for the headline claims is currently incomplete.

major comments (4)
  1. [Section III-C1, Eqs. (1), (2), (7)] The auxiliary loss in Eq. (7) directly minimizes the same Euclidean distance between Ec and T⊗Gc that Eq. (2) defines, while Eq. (1) combines the two quantities with attention A. If the auxiliary loss is effective, T⊗Gc converges toward Ec and the combined embedding in Eq. (1) becomes approximately Ec regardless of A, making the GSE pathway and the attention mechanism vacuous. If the auxiliary loss is weak—and the reported regularization parameter in Section IV-A3 is only 0.0001, presumably λ—then the paper's stated remedy for latent-space heterogeneity has no empirical support. The manuscript reports neither the achieved value of ||Ec−T⊗Gc||, nor a sensitivity analysis over λ, nor the learned attention values. This is load-bearing for the central claim that the transfer matrix addresses latent-space heterogeneity, and the current experiments do not reveal which of the two regimes holds.
  2. [Section IV-B, Tables II-IV] There is no source-free CDTM baseline. CDTM-H and CDTM differ from the DCN Base model not only in the use of source-domain data but also in architecture: they add the dual embedding structure, the combination attention, and the transfer matrix. Therefore the gains over Base cannot be attributed specifically to knowledge transfer from source domains rather than to the added capacity. To support the transfer claim, the authors should compare CDTM trained with target-domain data only against CDTM-H and CDTM under identical architecture and training settings.
  3. [Section IV-B, Tables II-V and Fig. 2] All offline results are reported as point estimates without standard errors, confidence intervals, or significance tests. This matters because several reported improvements are small; for example, on F2 in Table II the CDTM-H gain over DTCDR-H is 0.0028 AUC, and on F4 in Table III the CDTM-J gain over GA-DTCDR-J is 0.0005 AUC. The statement in Section IV-B3 that CDTM 'performs significantly' than CDTM-DA is not supported by any statistical test. The authors should provide error bars over multiple runs (or an equivalent uncertainty quantification) for the main comparisons.
  4. [Section IV-B3, Fig. 2 and Table V] The ablation study and the extensibility study are incomplete relative to the paper's stated contributions. The ablation removes only the combination attention (CDTM-DA), leaving the transfer matrix and the auxiliary loss unablated; contribution 3 in the introduction is therefore not tested. In Task 4, CDTM4 is compared only against the Base DCN model, not against the multi-domain CTR methods STAR and DADNN cited in Section II-B, so the claim that CDTM can be extended to multi-source and multi-target CDR is not benchmarked against existing multi-domain approaches.
minor comments (6)
  1. [Section III-B1] The first sentence says 'The transferable features have only one type of embedding (i.e., DSE), while the nontransferable features have two (DSE and GSE).' The following sentences state the opposite for both cases. The sentence appears to have transferable and nontransferable swapped and should be corrected, as the architecture description is otherwise inconsistent.
  2. [Eq. (2)] Eq. (2) writes L = argmin_T Σ ||Ec − T⊗Gc||²; this should be the minimum value of the objective or the minimizing argument, not the loss L equated to the argmin. The notation also conflicts with the use of L as a loss symbol elsewhere.
  3. [Table captions] The table captions are inconsistent with the task numbering: Table III is labeled 'TASK 2' but reports Task 1 using J as the source, and Table IV is labeled 'TASK 3' but reports the Task 2 multi-source comparison. Please align the captions with the task definitions in Section IV-A2.
  4. [Section III-C, Eq. (1)] The notation for T⊗Gc and Ec⊗A is not fully specified: A is called an attention matrix but described as a vector, and the dimensions of T relative to the embedding matrices are not stated. The paper should clarify whether ⊗ is element-wise multiplication and give the shapes of T and A.
  5. [Section IV-A1] The dataset description reports the number of transferable feature fields as 386 for H and 555 for J and F1-F4, but the text does not explain how the common transferable feature fields between each source-target pair are selected, nor how the differing feature fields are handled in the shared GSE table. A few sentences on this would improve reproducibility.
  6. [Throughout] There are several typographical errors, including 'clcik' in the Index Terms, 'domians' in Section II-B, and 'CTR estimation' in the abstract used where 'cross-domain recommendation' seems intended. A careful proofread is needed.

Circularity Check

1 steps flagged · score 4.0 of 10

The transfer-matrix 'alignment' is the training objective itself, so the claimed combination of distinct DSE/GSE information is enforced to collapse by construction, contradicting the paper's own 'should be different' premise.

  1. self definitional [Section III-C1 and Section III-E, Eqs. (1), (2), (7)]
    "To deal with the latent space heterogeneity of DSE and GSE, the transfer matrix is proposed to map them into a shared latent space. The optimization goal is to search for a transfer matrix that minimizes the Euclidean distance between the domain specific embedding Ec and global shared embedding Gc. L = argmin_T Σ ∥ Ec − T ⊗ Gc ∥^2 (2)... This transformation maintains two facts: (i) DSE and GSE should be different because they keep different information; (ii) the mapped GSE should be in the shared latent space with DSE. ..."

    Equation (7) is exactly the optimization objective of Eq. (2), so T is trained to make T⊗Gc approximate Ec on the target domain's own transferable features. In the combination Eq. (1), E = Ec⊗A + T⊗Gc⊗(1−A), the second term is therefore forced toward the first, making the two supposedly complementary embeddings duplicates rather than different sources of information. The paper even asserts fact (i) that DSE and GSE 'should be different', which the auxiliary loss directly contradicts. Because the paper never reports the residual distance, ablates λ, or shows that attention weights A are non-degenerate, the claimed benefit of the transfer-matrix alignment is definitional (enforced by the training objective) rather than empirically demonstrated.

full rationale

The paper's overall effectiveness claim is supported by external benchmarks (AUC comparisons against CoNet, DTCDR, GA-DTCDR, etc.) and an online A/B test, so the central empirical result is not circular. However, the specific mechanism claimed to solve latent-space heterogeneity—the transfer matrix—is self-definitional: its defining objective (Eq. 2) is inserted verbatim as the auxiliary loss (Eq. 7), so any 'alignment' achieved is enforced by construction rather than validated against an independent signal. Moreover, the paper's own stated fact that DSE and GSE 'should be different' is in direct tension with this objective, and the combination equation then combines two representations that the training loss drives to coincide. The paper provides no residual-distance measurement, no λ ablation, and no attention-weight analysis, so this mechanism-level circularity is real but partial: the multi-source and online gains could still come from other components or from the use of additional source data. Score 4 reflects one central mechanism reducing to its training objective while the overall system still has independent empirical content.

Assumptions & free parameters 3 free parameters · 3 assumptions · 0 invented entities

No external postulated entities such as new forces or particles appear in the paper. DSE, GSE, the transfer matrix, and attention are internal model components trained from data, not independent entities with falsifiable predictions outside the model. The main ledger items are the hand-set coefficients and the domain assumptions about feature transferability and linear alignability.

free parameters (3)
  • regularization coefficient lambda = 0.0001
    Set by hand in experiments; weights the auxiliary transfer-matrix loss in Eq. (7).
  • loss balance coefficients alpha_i = not reported
    Weights for combining domain losses and auxiliary losses in Eq. (8); values are not given in the paper.
  • embedding dimension k = not reported
    Dimension of DSE and GSE embeddings; chosen by hand but never stated in the paper.
assumptions (3)
  • domain assumption Transferable and non-transferable feature fields are known a priori for each domain.
    Section III-B defines X_c and X_d; the choice of which fields are transferable determines what is shared and is not learned or validated. If a supposedly transferable feature has conflicting semantics across domains, GSE would be polluted.
  • domain assumption Shared features across domains share a common feature-ID space and can be looked up in one global embedding table W_g.
    Section III-B assumes transferable features in different domains index into the same GSE table; the paper does not discuss ID alignment or semantic equivalence across domains.
  • domain assumption A per-domain linear transfer matrix T is sufficient to align the global shared embedding space with each domain-specific embedding space.
    Eq. (2) minimizes Euclidean distance between Ec and T times Gc; the paper assumes linear element-wise alignment resolves latent space heterogeneity without testing nonlinear alternatives.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Centralized-Distributed Transfer Model for Cross-Domain Recommendation Based on Multi-Source Heterogeneous Transfer Learning." pith.science (2026). https://pith.science/paper/2GDWV7HY

@misc{pith2026241109286,
  author       = {Pith},
  title        = {Pith review of: A Centralized-Distributed Transfer Model for Cross-Domain Recommendation Based on Multi-Source Heterogeneous Transfer Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2GDWV7HY}},
  note         = {Machine review of arXiv:2411.09286}
}
read the original abstract

Cross-domain recommendation (CDR) methods are proposed to tackle the sparsity problem in click through rate (CTR) estimation. Existing CDR methods directly transfer knowledge from the source domains to the target domain and ignore the heterogeneities among domains, including feature dimensional heterogeneity and latent space heterogeneity, which may lead to negative transfer. Besides, most of the existing methods are based on single-source transfer, which cannot simultaneously utilize knowledge from multiple source domains to further improve the model performance in the target domain. In this paper, we propose a centralized-distributed transfer model (CDTM) for CDR based on multi-source heterogeneous transfer learning. To address the issue of feature dimension heterogeneity, we build a dual embedding structure: domain specific embedding (DSE) and global shared embedding (GSE) to model the feature representation in the single domain and the commonalities in the global space,separately. To solve the latent space heterogeneity, the transfer matrix and attention mechanism are used to map and combine DSE and GSE adaptively. Extensive offline and online experiments demonstrate the effectiveness of our model.

Figures

Figures reproduced from arXiv: 2411.09286 by the authors.

Figure 1
Figure 1. The Framework of proposed Centralized-Distributed Transfer Model (GSE ie. Global shared embedding, DSE ie. Domain specific embedding. The [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Comparison of ablation experiments (DA represents the model that [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

28 extracted references · 22 canonical work pages

  1. [1]

    Deepfm: a factorization- machine based neural network for ctr prediction,

    H. Guo, R. Tang, Y . Ye, Z. Li, and X. He, “Deepfm: a factorization- machine based neural network for ctr prediction,” arXiv preprint arXiv:1703.04247, 2017

    arXiv 2017

  2. [2]

    Deep interest network for click-through rate prediction,

    G. Zhou, X. Zhu, C. Song, Y . Fan, H. Zhu, X. Ma, Y . Yan, J. Jin, H. Li, and K. Gai, “Deep interest network for click-through rate prediction,” in Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining , 2018, pp. 1059–1068

    work page 2018

  3. [3]

    A Survey on Cross-domain Recommendation: Taxonomies, Methods, and Future Directions

    T. Zang, Y . Zhu, H. Liu, R. Zhang, and J. Yu, “A survey on cross- domain recommendation: Taxonomies, methods, and future directions,” arXiv preprint arXiv:2108.03357 , 2021

    work page Pith review arXiv 2021

  4. [4]

    Predicting clicks: estimating the click-through rate for new ads,

    M. Richardson, E. Dominowska, and R. Ragno, “Predicting clicks: estimating the click-through rate for new ads,” in Proceedings of the 16th international conference on World Wide Web , 2007, pp. 521–530

    work page 2007

  5. [5]

    Factorization machines,

    S. Rendle, “Factorization machines,” in 2010 IEEE International con- ference on data mining . IEEE, 2010, pp. 995–1000

    2010

  6. [6]

    Field-aware factorization machines for ctr prediction,

    Y . Juan, Y . Zhuang, W.-S. Chin, and C.-J. Lin, “Field-aware factorization machines for ctr prediction,” in Proceedings of the 10th ACM conference on recommender systems , 2016, pp. 43–50

    work page 2016

  7. [7]

    Product- based neural networks for user response prediction,

    Y . Qu, H. Cai, K. Ren, W. Zhang, Y . Yu, Y . Wen, and J. Wang, “Product- based neural networks for user response prediction,” in 2016 IEEE 16th International Conference on Data Mining (ICDM) . IEEE, 2016, pp. 1149–1154

    work page 2016

  8. [8]

    Wide & deep learning for recommender systems,

    H.-T. Cheng, L. Koc, J. Harmsen, T. Shaked, T. Chandra, H. Aradhye, G. Anderson, G. Corrado, W. Chai, M. Ispir et al. , “Wide & deep learning for recommender systems,” in Proceedings of the 1st workshop on deep learning for recommender systems , 2016, pp. 7–10

    work page 2016

Show all 28 references
  1. [9]

    Deep & cross network for ad click predictions,

    R. Wang, B. Fu, G. Fu, and M. Wang, “Deep & cross network for ad click predictions,” in Proceedings of the ADKDD’17 , 2017, pp. 1–7

    2017

  2. [10]

    Adaptive factorization network: Learning adaptive-order feature interactions,

    W. Cheng, Y . Shen, and L. Huang, “Adaptive factorization network: Learning adaptive-order feature interactions,” inProceedings of the AAAI Conference on Artificial Intelligence , vol. 34, no. 04, 2020, pp. 3609– 3616

    2020

  3. [11]

    Dcn v2: Improved deep & cross network and practical lessons for web- scale learning to rank systems,

    R. Wang, R. Shivanna, D. Cheng, S. Jain, D. Lin, L. Hong, and E. Chi, “Dcn v2: Improved deep & cross network and practical lessons for web- scale learning to rank systems,” in Proceedings of the Web Conference 2021, 2021, pp. 1785–1797

    2021

  4. [12]

    Deep interest evolution network for click-through rate prediction,

    G. Zhou, N. Mou, Y . Fan, Q. Pi, W. Bian, C. Zhou, X. Zhu, and K. Gai, “Deep interest evolution network for click-through rate prediction,” in Proceedings of the AAAI conference on artificial intelligence , vol. 33, no. 01, 2019, pp. 5941–5948

    2019

  5. [13]

    Multi-interest network with dynamic routing for recommendation at tmall,

    C. Li, Z. Liu, M. Wu, Y . Xu, H. Zhao, P. Huang, G. Kang, Q. Chen, W. Li, and D. L. Lee, “Multi-interest network with dynamic routing for recommendation at tmall,” in Proceedings of the 28th ACM international conference on information and knowledge management, 2019, pp. 2615– 2623

    2019

  6. [14]

    Practice on long sequential user behavior modeling for click-through rate prediction,

    Q. Pi, W. Bian, G. Zhou, X. Zhu, and K. Gai, “Practice on long sequential user behavior modeling for click-through rate prediction,” in Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining , 2019, pp. 2671–2679

    2019

  7. [15]

    Search-based user interest modeling with lifelong sequential behavior data for click-through rate prediction,

    Q. Pi, G. Zhou, Y . Zhang, Z. Wang, L. Ren, Y . Fan, X. Zhu, and K. Gai, “Search-based user interest modeling with lifelong sequential behavior data for click-through rate prediction,” in Proceedings of the 29th ACM International Conference on Information & Knowledge Managemen...

    2020

  8. [16]

    Cross-domain recommendation: An embedding and mapping approach

    T. Man, H. Shen, X. Jin, and X. Cheng, “Cross-domain recommendation: An embedding and mapping approach.” in IJCAI, vol. 17, 2017, pp. 2464–2470

    2017

  9. [17]

    Dtcdr: A framework for dual-target cross-domain recommendation,

    F. Zhu, C. Chen, Y . Wang, G. Liu, and X. Zheng, “Dtcdr: A framework for dual-target cross-domain recommendation,” in Proceedings of the 28th ACM International Conference on Information and Knowledge Management, 2019, pp. 1533–1542

    2019

  10. [18]

    Cross-domain attention network with wasserstein regularizers for e-commerce search,

    M. Qiu, B. Wang, C. Chen, X. Zeng, J. Huang, D. Cai, J. Zhou, and F. S. Bao, “Cross-domain attention network with wasserstein regularizers for e-commerce search,” in Proceedings of the 28th ACM International Conference on Information and Knowledge Management , 2019, pp. 2509–2515

    2019

  11. [19]

    A deep framework for cross-domain and cross-system recommendations,

    F. Zhu, Y . Wang, C. Chen, G. Liu, M. Orgun, and J. Wu, “A deep framework for cross-domain and cross-system recommendations,” arXiv preprint arXiv:2009.06215, 2020

    2009 arXiv

  12. [20]

    A unified framework for cross-domain and cross-system recommendations,

    F. Zhu, Y . Wang, J. Zhou, C. Chen, L. Li, and G. Liu, “A unified framework for cross-domain and cross-system recommendations,” IEEE Transactions on Knowledge and Data Engineering , 2021

    2021

  13. [21]

    Darec: Deep domain adaptation for cross-domain recommendation via transferring rating patterns,

    F. Yuan, L. Yao, and B. Benatallah, “Darec: Deep domain adaptation for cross-domain recommendation via transferring rating patterns,” arXiv preprint arXiv:1905.10760, 2019

    1905 arXiv

  14. [22]

    Conet: Collaborative cross networks for cross-domain recommendation,

    G. Hu, Y . Zhang, and Q. Yang, “Conet: Collaborative cross networks for cross-domain recommendation,” in Proceedings of the 27th ACM international conference on information and knowledge management , 2018, pp. 667–676

    2018

  15. [23]

    Ddtcdr: Deep dual transfer cross domain recommendation,

    P. Li and A. Tuzhilin, “Ddtcdr: Deep dual transfer cross domain recommendation,” in Proceedings of the 13th International Conference on Web Search and Data Mining , 2020, pp. 331–339

    2020

  16. [24]

    Minet: Mixed interest network for cross-domain click- through rate prediction,

    W. Ouyang, X. Zhang, L. Zhao, J. Luo, Y . Zhang, H. Zou, Z. Liu, and Y . Du, “Minet: Mixed interest network for cross-domain click- through rate prediction,” in Proceedings of the 29th ACM International Conference on Information & Knowledge Management, 2020, pp. 2669– 2676

    2020

  17. [25]

    Dual attentive sequential learning for cross-domain click-through rate prediction,

    P. Li, Z. Jiang, M. Que, Y . Hu, and A. Tuzhilin, “Dual attentive sequential learning for cross-domain click-through rate prediction,” arXiv preprint arXiv:2106.02768 , 2021

    2021 arXiv

  18. [26]

    Personalized transfer of user preferences for cross-domain recommendation,

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

    2022

  19. [27]

    Dadnn: Multi- scene ctr prediction via domain-aware deep neural network,

    J. He, G. Mei, F. Xing, X. Yang, Y . Bao, and W. Yan, “Dadnn: Multi- scene ctr prediction via domain-aware deep neural network,” arXiv preprint arXiv:2011.11938, 2020

    2011 arXiv

  20. [28]

    One model to serve all: Star topology adaptive recommender for multi-domain ctr prediction,

    X.-R. Sheng, L. Zhao, G. Zhou, X. Ding, B. Dai, Q. Luo, S. Yang, J. Lv, C. Zhang, H. Deng et al., “One model to serve all: Star topology adaptive recommender for multi-domain ctr prediction,” in Proceedings of the 30th ACM International Conference on Information & Knowledge Ma...

    2021

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.