Pith. sign in

REVIEW 4 major objections 4 minor 2 cited by

Controlling Diversity at Inference: Guiding Diffusion Recommender Models with Targeted Category Preferences

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

Pith's one-line read A category-preference vector fed into each denoising step lets one diffusion recommender tune the accuracy-diversity trade-off at inference time and follow arbitrary target category mixes without retraining.

desk verdict D3Rec makes a credible extension of DiffRec for inference-time diversity control, but the 'arbitrary preferences' claim hinges on a semi-synthetic experiment run at w=7, outside the reported tuning grid, with no error bars. read the letter →

arxiv 2411.11240 v2 pith:EKXBHKJV submitted 2024-11-18 cs.IR

classification cs.IR
keywords D3Recinference-timediversitycontroldiffusionrecommendercategorypreferenceconditioningclassifier-freeguidancedisentangledrepresentationaccuracy-diversitytrade-offfilterbubble
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

Recommender systems usually settle the accuracy-diversity trade-off during training, so changing how diverse recommendations should be means retraining or waiting for new user feedback. This paper proposes D3Rec, a diffusion-based recommender that instead treats diversity as an inference-time input: a category-preference vector is fed into every denoising step of the generative process. At inference, that vector can be smoothed with a temperature applied to the user's own preference distribution, which moves recommendations along the accuracy-diversity frontier, or replaced with an arbitrary targeted vector, for instance shifting a movie-watcher from action to romance. Across three real-world datasets and semi-synthetic sets, the authors report that a single trained model improves over the best of ten baselines on almost all accuracy and diversity metrics and adapts quickly to category mixes very different from the user's history. If the paper is right, D3Rec turns diversity into a runtime dial rather than a training-time hyperparameter, which matters for filter-bubble mitigation and fast-changing business strategies.

What carries the argument

The load-bearing object is the targeted category-preference vector $\tilde{\mathbf y}\in\mathbb{R}^{|C|}$, normalized to sum to 1, which serves both as the conditioning input of the reverse diffusion process and as the inference-time control knob. The mechanism that lets the model follow arbitrary targets is classifier-free guidance applied to the predicted denoised interactions, $\tilde{\mathbf x}_\theta(\mathbf x_t,t,\tilde{\mathbf y})=(1+w)\,\mathbf x_\theta(\mathbf x_t,t,\tilde{\mathbf y})-w\,\mathbf x_\theta(\mathbf x_t,t,\mathbf 0_{|C|})$, where $w$ is the guidance strength and the unconditional term is the same denoiser evaluated on a zero condition. The architecture wraps this in a U-Net-like decoder and two-tower encoders: a category-aware tower that receives projected category embeddings at every layer, a category-independent tower, and an orthogonal disentanglement loss that keeps the two representations separate. The temperature transform $\tilde{\mathbf y}=\mathrm{Softmax}(\log \mathbf y/\tau)$ converts the model into the paper's smooth diversity dial, while direct substitution of $\tilde{\mathbf y}$ handles arbitrary targets.

What would settle it

An experiment could feed D3Rec targeted category vectors of the form $\tilde{\mathbf y}=(1-\lambda)\mathbf y+\lambda\mathbf e_c$ for $\lambda\in[0,1]$ and each category $c$, sweeping the guidance strength over the full grid plus $w=7$, and check whether Entropy@K moves monotonically with $\lambda$ while the category mix of the final top-K list matches $\tilde{\mathbf y}$. A mismatch or collapse at intermediate mixes outside the training support would falsify the arbitrary-target claim.

Watch

Extended reading notes

Core claim

The central claim is that category preference can be made a controllable condition of diffusion-based recommendation. D3Rec first runs the forward diffusion process, adding Gaussian noise to a user's interaction vector so that the category preferences lurking in the history are erased; the reverse process then regenerates the future interaction vector under the targeted category preference $\tilde{\mathbf y}$, using a denoiser whose encoder and decoder each inject $\tilde{\mathbf y}$ through learned category embeddings. To make the generated lists actually match the target, the paper adds two auxiliary losses, one that predicts the category mix of the reconstructed history and one that associates category embeddings with items belonging to each category, plus a re-weighting that reduces the gradient dominance of popular categories. At inference, $\tilde{\mathbf y}$ can be obtained by temperature-smoothing the user's own preference vector or by setting any arbitrary vector summing to 1, and classifier-free guidance with strength $w$ controls how strongly the target steers the denoising. On this basis the paper reports that a single trained model moves entropy and coverage monotonically with temperature while keeping accuracy, and that its accuracy-diversity Pareto frontier dominates the compared baselines on the three datasets.

Load-bearing premise

The load-bearing premise is that the model, trained only with the target category preference equal to the user's original preference, will still steer generation correctly toward arbitrary category preferences far outside what it saw in training; the paper's strongest evidence for this is a semi-synthetic experiment run at a strength setting outside the range used in its main experiments.

Editorial extensions

If this is right

  • A deployed model can serve different diversity levels to different users or on different days without retraining: the same checkpoint answers to whichever $\tilde{\mathbf y}$ is fed in.
  • The Pareto curves imply that at any given accuracy level D3Rec can produce more diverse lists than the compared end-to-end and post-processing methods, and at any given diversity it loses less accuracy.
  • The semi-synthetic results imply that the conditioning mechanism transfers to category preferences far from the user's own history, so the system can react to short-term shifts in taste, such as a request for a genre the user rarely watches.
  • Because inference is the same denoising loop as the base diffusion recommender, the added controllability costs little in latency compared with re-ranking methods.

Reading between the lines

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

  • Inference: a natural extension the paper does not test is to set the target vector as a convex combination of the user's preference and a business-specified category mix; the monotonic temperature curves suggest coverage and entropy should interpolate smoothly, giving an operational rule for live A/B tests.
  • Inference: the same conditioning-by-attribute design could apply to item attributes beyond category, such as price tier, provider, or popularity band, so the mechanism might generalize from diversity control to fairness-style constraints.
  • Inference: the semi-synthetic success at guidance strength $w=7$, outside the grid searched for the real-data experiments, leaves open that optimal strength grows with the distance between training condition and target; this is a testable prediction, not a result the paper establishes.
  • Inference: because the forward process deliberately erases category signals, D3Rec may also reduce bias amplification in calibrated-recommendation settings, but the paper itself measures accuracy, diversity, and noise stability, not calibration.
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 / 4 minor

Summary. The paper proposes D3Rec, a diffusion-based recommender that conditions generation on a category-preference vector. The model uses disentangled two-tower encoders, classifier-free guidance, an auxiliary category-matching loss, and a focal-style re-weighting strategy. At inference, the category condition can be modified by a temperature parameter or set to an arbitrary target vector, with the aim of controlling the accuracy-diversity trade-off without retraining. The paper evaluates D3Rec on ML-1M, Steam, and Anime 2023, reporting Pareto curves, a semi-synthetic adaptation experiment, ablations, noise robustness, and inference latency.

Significance. If the claims hold, the contribution is practically valuable: it moves the accuracy-diversity trade-off to inference time, which existing end-to-end diversity methods do not offer. The paper is generally well structured, includes multiple datasets and baselines, reports significance tests for the main table, and provides an anonymized code link. The main experimental evidence for the accuracy-diversity frontier (Table 2 and Figure 3) is reasonably strong. The weakest and most load-bearing claim is the third desideratum, 'adapting to arbitrary targeted category preferences': the only direct evidence is a semi-synthetic experiment with an unexplained choice of guidance strength (w=7) outside the reported search grid, and the table reports no variance. The central derivation is otherwise internally consistent, so the paper is close, but the arbitrary-preference claim needs stronger support before acceptance.

major comments (4)
  1. [Section 5.4, Table 4] The semi-synthetic experiment that is the only direct evidence for the 'arbitrary targeted category preferences' claim fixes the guidance strength at w=7, whereas the hyperparameter grids in Section 5.1 and Appendix A are restricted to {-0.7,-0.5,-0.3,0,0.3,0.5}. No justification is given for w=7, and if it was selected using the same semi-synthetic test distributions, the evaluation is not a clean out-of-distribution test. Moreover, Table 4 reports no standard deviations or significance tests, unlike Table 2, so it is impossible to assess the stability of the large gains (e.g., +143% Recall@10 on ML-1M). Please report results for validation-tuned w (e.g., by holding out a subset of users or target vectors) and include variance across seeds.
  2. [Section 4.4.1, Eq. (12)] The re-weight strategy divides by max(1-y)-min(1-y) and max(y)-min(y); for users whose category preference vector is uniform or has all entries equal, this denominator is zero and the weights become undefined. Since the paper does not state a fallback, this edge case can break training on real data. Add an epsilon or a uniform-weight fallback and describe its use.
  3. [Section 4.1.2 / Section 4.3.1] The model is trained only with the condition y equal to the user's original category preference, plus condition dropout, so the ability to follow an arbitrary y_tilde at inference is an extrapolation of classifier-free guidance in Eq. (6). The paper should provide a more direct test of this extrapolation, e.g., by sampling target vectors across the simplex or using held-out target distributions, and should ablate the auxiliary loss L_cate to show that the guidance, not merely the training objective, generalizes to unseen category mixes.
  4. [Section 5.4] The semi-synthetic test construction uses only the bottom 30% categories per user as the test target. This is a specific and somewhat extreme shift, not a demonstration of 'arbitrary' preferences. A small grid of target vectors (e.g., uniform, category-shifted, and interpolated targets) would make the claim much more convincing and would also help justify the choice of w.
minor comments (4)
  1. [Section 5.1 vs Appendix A] The diffusion step grid is given as {10,15,20,100} in Section 5.1 and {5,15,40,100} in Appendix A; Table 7 then uses {10,15,20,100}. Please reconcile these inconsistencies.
  2. [Table 4] For Anime 2023, the MultVAE NDCG@10 value 0.00129 appears to be a typo (likely 0.0129), since it is an order of magnitude lower than the corresponding NDCG@20 value.
  3. [Section 4.4.1] The text says 'loss weight upper bound gamma_min', but gamma_min is defined as the lower bound. Please correct the wording.
  4. [Section 5.5 / Figure 4] The inference-time comparison would be more informative if the reported values included absolute times or at least a unit; as written, only relative ordering is visible.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: D3Rec's inference-time diversity control is an explicit training objective, and the paper's generalization and trade-off claims are tested against held-out data and baselines.

full rationale

D3Rec's inference-time control is explicitly engineered rather than discovered: the model is conditioned on the targeted category preference y~ (Eqs. 5-6) and trained with the auxiliary loss Lcate (Eq. 10) so that the category distribution of reconstructions matches y~. Since Entropy@K and Coverage@K (Eq. 15) are computed precisely from the category distribution of the top-K list, the observed monotonic diversity change with temperature is a direct consequence of the training objective. This is the proposed mechanism, not a hidden prediction, and the paper does not present it as an empirical discovery. The load-bearing claims—that D3Rec improves the accuracy-diversity frontier over ten baselines (Table 2, Figure 3) and that it can follow arbitrary targeted preferences outside training support (Table 4, semi-synthetic)—are evaluated against held-out interactions and baselines, giving independent content. The only self-citations ([10]-[12]) are background references on recommender systems and are not load-bearing. A methodological caveat is that Section 5.4 fixes the guiding strength w at 7 for the semi-synthetic experiment although the grid in Section 5.1 for w is {-0.7,-0.5,-0.3,0,0.3,0.5}, and Table 4 reports no variance; this is a validation/leakage concern for the generalization claim, not a definitional circularity, and should be weighted under correctness risk. Overall, no step in the derivation chain reduces to its own input by construction.

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

The central design depends on several standard ML ingredients and a few ad-hoc components. The most consequential free parameter is w, since the headline adaptation result uses a value outside the grid. The category manipulations rely on F being accurate, and the disentanglement and guidance assumptions are asserted rather than demonstrated.

free parameters (4)
  • w (classifier-free guidance strength) = 0.5 (grid max); 7 (synthetic)
    Eq. 6 balance between conditional and unconditional prediction; w=7 used for arbitrary-preference experiment is outside reported grid.
  • lambda = 1, 1e-2, or 1e-4
    Weight for Lortho + Lemb in Eq. 14.
  • gamma_min, gamma_max = 0.3-1 and 1-2
    Bounds for re-weight vectors y_pos and y_neg in Section 4.4.1.
  • diffusion steps T = 10-100, best 15
    Number of denoising steps; tuned per dataset (Table 7).
assumptions (4)
  • domain assumption DDPM Gaussian forward/reverse process is a valid model for binarized interaction vectors
    Preliminary and Section 4.1 treat x0 in {0,1}^|I| as if corrupted by Gaussian noise; diffusion models for discrete data typically require continuous embeddings or multinomial formulations.
  • domain assumption The item-category matrix F is complete and correctly represents item categories
    Category preference y_u and losses Lcate/Lemb are computed from F; errors in F would propagate into the condition and the evaluation.
  • ad hoc to paper Cosine loss Lortho enforces meaningful disentanglement between category and non-category features
    Eq. 8 penalizes cosine similarity, but there is no quantitative check that the category-independent encoder truly excludes category information.
  • domain assumption Classifier-free guidance transfers from continuous image generation to discrete interaction vectors
    Eq. 6 is borrowed from Ho and Salimans [8]; the paper does not justify its validity on binary high-dimensional interaction data.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Controlling Diversity at Inference: Guiding Diffusion Recommender Models with Targeted Category Preferences." pith.science (2026). https://pith.science/paper/EKXBHKJV

@misc{pith2026241111240,
  author       = {Pith},
  title        = {Pith review of: Controlling Diversity at Inference: Guiding Diffusion Recommender Models with Targeted Category Preferences},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EKXBHKJV}},
  note         = {Machine review of arXiv:2411.11240}
}
read the original abstract

Diversity control is an important task to alleviate bias amplification and filter bubble problems. The desired degree of diversity may fluctuate based on users' daily moods or business strategies. However, existing methods for controlling diversity often lack flexibility, as diversity is decided during training and cannot be easily modified during inference. We propose \textbf{D3Rec} (\underline{D}isentangled \underline{D}iffusion model for \underline{D}iversified \underline{Rec}ommendation), an end-to-end method that controls the accuracy-diversity trade-off at inference. D3Rec meets our three desiderata by (1) generating recommendations based on category preferences, (2) controlling category preferences during the inference phase, and (3) adapting to arbitrary targeted category preferences. In the forward process, D3Rec removes category preferences lurking in user interactions by adding noises. Then, in the reverse process, D3Rec generates recommendations through denoising steps while reflecting desired category preferences. Extensive experiments on real-world and synthetic datasets validate the effectiveness of D3Rec in controlling diversity at inference.

Figures

Figures reproduced from arXiv: 2411.11240 by the authors.

Figure 1
Figure 1. Diversity control: (a) Existing methods [ [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The overall framework of D3Rec. In the forward process, the user interactions are corrupted, thereby diminishing [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Accuracy-diversity curves on three real-world datasets. The closer to the top right corner, the better the trade-off [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Inference time comparison. mismatch in category preferences between the training dataset and the test dataset. The data statistics for three semi-synthetic datasets are presented in [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: Effect of the guiding strength 𝑤 on real-world (left) and synthetic datasets (right) [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

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

  1. Collaborative Diffusion Model for Recommender System

    cs.IR 2025-01 conditional novelty 5.0 of 10

    CDiff4Rec improves diffusion recommenders by injecting item-content pseudo-users and real-user neighbor predictions into the denoising objective, beating DiffRec and other baselines on Yelp, Amazon-Game, and Citeulike-t.

  2. Multi-Objective Recommendation in the Era of Generative AI: A Survey of Recent Progress and Future Prospects

    cs.IR 2025-06 conditional novelty 4.0 of 10

    A survey classifying multi-objective recommendation research that uses generative AI into objective families, with metrics, datasets, and future directions.

Reference graph

Works this paper leans on

45 extracted references · 37 canonical work pages · cited by 2 Pith papers

  1. [1]

    Shilong Bao, Qianqian Xu, Zhiyong Yang, Yuan He, Xiaochun Cao, and Qingming Huang. 2022. The minority matters: A diversity-promoting collaborative metric learning algorithm. In NeurIPS

  2. [2]

    Jaime Carbonell and Jade Goldstein. 1998. The use of MMR, diversity-based reranking for reordering documents and producing summaries. In SIGIR

  3. [3]

    Yukuo Cen, Jianwei Zhang, Xu Zou, Chang Zhou, Hongxia Yang, and Jie Tang

  4. [4]

    Jaehoon Cha and Jeyan Thiyagalingam. 2023. Orthogonality-enforced latent space in autoencoders: An approach to learning disentangled representations. In ICML

  5. [5]

    Laming Chen, Guoxin Zhang, and Eric Zhou. 2018. Fast greedy map inference for determinantal point process to improve recommendation diversity. In NeurIPS

  6. [6]

    Guibing Guo, Huan Zhou, Bowei Chen, Zhirong Liu, Xiao Xu, Xu Chen, Zhenhua Dong, and Xiuqiang He. 2020. IPGAN: Generating informative item pairs by adversarial sampling. IEEE transactions on neural networks and learning systems (2020)

  7. [7]

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. 2020. Denoising diffusion probabilistic models. In NeurIPS

  8. [8]

    Jonathan Ho and Tim Salimans. 2021. Classifier-Free Diffusion Guidance. In NeurIPS Workshop on Deep Generative Models and Downstream Applications

Show all 45 references
  1. [9]

    Diederik P Kingma and Max Welling. 2014. Auto-encoding variational bayes. In ICLR

  2. [10]

    Wonbin Kweon, SeongKu Kang, Sanghwan Jang, and Hwanjo Yu. 2024. Top- Personalized-K Recommendation. In WWW

  3. [11]

    Wonbin Kweon and Hwanjo Yu. 2024. Doubly Calibrated Estimator for Recom- mendation on Data Missing Not At Random. In WWW

  4. [12]

    Gyuseok Lee, SeongKu Kang, Wonbin Kweon, and Hwanjo Yu. 2024. Continual Collaborative Distillation for Recommender System. In KDD

  5. [13]

    Xiang Li, John Thickstun, Ishaan Gulrajani, Percy S Liang, and Tatsunori B Hashimoto. 2022. Diffusion-lm improves controllable text generation. InNeurIPS

  6. [14]

    Dawen Liang, Rahul G Krishnan, Matthew D Hoffman, and Tony Jebara. 2018. Variational autoencoders for collaborative filtering. In WWW

  7. [15]

    Tsung-Yi Lin, Priya Goyal, Ross Girshick, Kaiming He, and Piotr Dollár. 2017. Focal loss for dense object detection. In ICCV

  8. [16]

    Zihan Lin, Hui Wang, Jingshu Mao, Wayne Xin Zhao, Cheng Wang, Peng Jiang, and Ji-Rong Wen. 2022. Feature-aware diversified re-ranking with disentangled representations for relevant recommendation. In KDD

  9. [17]

    Ilya Loshchilov and Frank Hutter. 2019. Decoupled weight decay regularization. In ICLR

  10. [18]

    Calvin Luo. 2022. Understanding diffusion models: A unified perspective. arXiv preprint arXiv:2208.11970 (2022)

  11. [19]

    Jianxin Ma, Chang Zhou, Peng Cui, Hongxia Yang, and Wenwu Zhu. 2019. Learn- ing disentangled representations for recommendation. In NeurIPS

  12. [20]

    Sayooran Nagulendra and Julita Vassileva. 2014. Understanding and controlling the filter bubble through interactive visualization: a user study. In HT

  13. [21]

    Apurva Pathak, Kshitiz Gupta, and Julian McAuley. 2017. Generating and per- sonalizing bundle recommendations on steam. In SIGIR

  14. [22]

    Lijing Qin and Xiaoyan Zhu. 2013. Promoting diversity in recommendation by entropy regularizer. In IJCAI

  15. [23]

    Yuanyi Ren, Hang Ni, Yingxue Zhang, Xi Wang, Guojie Song, Dong Li, and Jianye Hao. 2023. Dual-Process Graph Neural Network for Diversified Recommendation. In CIKM

  16. [24]

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Björn Ommer. 2022. High-resolution image synthesis with latent diffusion models. In CVPR

  17. [25]

    Olaf Ronneberger, Philipp Fischer, and Thomas Brox. 2015. U-net: Convolutional networks for biomedical image segmentation. In MICCAI

  18. [26]

    Chaofeng Sha, Xiaowei Wu, and Junyu Niu. 2016. A framework for recommending relevant and diverse items.. In IJCAI

  19. [27]

    Harald Steck. 2018. Calibrated recommendations. In RecSys

  20. [28]

    Ewout Van Den Berg and Michael P Friedlander. 2009. Probing the Pareto frontier for basis pursuit solutions. Siam journal on scientific computing 31, 2 (2009), 890– 912

  21. [29]

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. In NeurIPS. KDD ’25, August 03 – 07, 2025, Toronto, ON, Canada Gwangseok Han*, Wonbin Kweon*, Minsoo Kim, and Hwanjo Yu

  22. [30]

    Wenjie Wang, Fuli Feng, Xiangnan He, Xiang Wang, and Tat-Seng Chua. 2021. Deconfounded recommendation for alleviating bias amplification. In KDD

  23. [31]

    Wenjie Wang, Fuli Feng, Liqiang Nie, and Tat-Seng Chua. 2022. User-controllable recommendation against filter bubbles. In SIGIR

  24. [32]

    Wenjie Wang, Xinyu Lin, Fuli Feng, Xiangnan He, Min Lin, and Tat-Seng Chua

  25. [33]

    Wenjie Wang, Yiyan Xu, Fuli Feng, Xinyu Lin, Xiangnan He, and Tat-Seng Chua

  26. [34]

    Qiong Wu, Yong Liu, Chunyan Miao, Binqiang Zhao, Yin Zhao, and Lu Guan

  27. [35]

    Liangwei Yang, Shengjie Wang, Yunzhe Tao, Jiankai Sun, Xiaolong Liu, Philip S Yu, and Taiqing Wang. 2023. DGRec: Graph Neural Network for Recommendation with Diversified Embedding Generation. In WSDM

  28. [36]

    Zhengyi Yang, Jiancan Wu, Zhicai Wang, Xiang Wang, Yancheng Yuan, and Xiangnan He. 2024. Generate What You Prefer: Reshaping Sequential Recom- mendation via Guided Diffusion. In NeurIPS

  29. [37]

    Xiaoying Zhang, Hongning Wang, and Hang Li. 2023. Disentangled Representa- tion for Diversified Recommendations. In WSDM

  30. [38]

    Yuying Zhao, Minghua Xu, Huiyuan Chen, Yuzhong Chen, Yiwei Cai, Rashidul Islam, Yu Wang, and Tyler Derr. 2024. Can One Embedding Fit All? A Multi- Interest Learning Paradigm Towards Improving User Interest Diversity Fairness. In WWW

  31. [39]

    Yu Zheng, Chen Gao, Liang Chen, Depeng Jin, and Yong Li. 2021. DGCN: Diver- sified recommendation with graph convolutional networks. In WWW

  32. [40]

    Tao Zhou, Zoltán Kuscsik, Jian-Guo Liu, Matúš Medo, Joseph Rushton Wakeling, and Yi-Cheng Zhang. 2010. Solving the apparent diversity-accuracy dilemma of recommender systems. Proceedings of the National Academy of Sciences (2010)

  33. [41]

    Cai-Nicolas Ziegler, Sean M McNee, Joseph A Konstan, and Georg Lausen. 2005. Improving recommendation lists through topic diversification. In WWW. Controlling Diversity at Inference: Guiding Diffusion Recommender Models with Targeted Category Preferences KDD ’25, August 03 – 0...

  34. [2019]

    In IJCAI

    PD-GAN: Adversarial learning for personalized diversity-promoting rec- ommendation.. In IJCAI

  35. [2020]

    Controllable multi-interest framework for recommendation. In KDD

  36. [2022]

    Causal representation learning for out-of-distribution recommendation. In WWW

  37. [2023]

    In SIGIR

    Diffusion Recommender Model. In SIGIR

Pith tools

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