Equity by Design? On the Trade-Offs in Fairness-Driven Recommendation in Heterogeneous Two-Sided Markets

Alexander Timans; Dominykas Seputis; Rajeev Verma

REVIEW 5 major objections 4 minor 36 references

Equity by Design? On the Trade-Offs in Fairness-Driven Recommendation in Heterogeneous Two-Sided Markets

T0 review · 5 major / 4 minor · reviewed 2026-08-03 · deepseek-v4-flash

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

Pith's one-line read The 'free fairness' regime for producer exposure in recommendation systems survives only in single-item settings; with multi-item lists it vanishes, and moderate fairness constraints can instead boost business metrics.

desk verdict The paper's main empirical claim — that 'free fairness' doesn't survive multi-item lists — is plausible and worth testing, but the evidence as reported is shaky: the GMV constraint is vacuous, the recommender's near-perfect precision suggests training-set evaluation, and the business metric gains may be an artifact of the purchase simulation. read the letter →

arxiv 2602.10739 v3 pith:BID4OK5Q submitted 2026-02-11 cs.GT cs.IR

Dominykas Seputis , Alexander Timans , Rajeev Verma This is my paper

classification cs.GTcs.IR

keywords two-sidedmarketsrecommendersystemsfairnessfreeCVaRproducerexposurebusinessconstraintsmulti-itemrecommendation

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 challenges a widely held result in two-sided recommender design: that producer fairness (guaranteeing minimum exposure) can be free for consumers. It argues this 'free fairness' regime was an artifact of single-item recommendations (k=1) with many consumers per producer. In the realistic discrete multi-item setting (k>1), enforcing producer exposure cuts mean consumer utility by 15–25% at k=10, depending on dataset. However, moderate producer fairness (gamma ~0.3–0.6) can raise simulated sell-through rate and gross merchandise value by diversifying exposure away from saturated producers. The paper also introduces a CVaR objective over consumer groups that compresses inter-group utility disparities without meaningful degradation elsewhere.

What carries the argument

A discrete multi-item allocation formulation (binary w_ij, exactly k items per consumer) coupled with a Conditional Value-at-Risk objective over consumer groups and a GMV threshold constraint. The binary multi-item structure is the mechanism that breaks 'free fairness': with k>1 each consumer's list must go beyond top choices, creating coupling and relevance cost that soft single-item allocations did not capture.

What would settle it

Re-run the allocation framework with a purchase model where consumers make at most one purchase per session or producers have inventory that does not automatically sell out; if moderate fairness no longer raises STR/GMV, the business-alignment conclusion collapses.

Watch

Extended reading notes

Core claim

The paper's central claim is that the 'free fairness' regime—where producer exposure constraints impose no consumer cost—holds only in the stylized single-item, consumer-heavy setting. Once recommendations become discrete multi-item lists (each consumer receives k items, binary allocations), the regime disappears: at k=10 with full producer fairness, mean consumer utility drops 15–25% across three datasets. The paper further shows that moderate producer constraints can improve simulated business metrics, that CVaR group fairness reduces tail-group disparities, and that scalable solvers (LP relaxation, augmented Lagrangian) match exact mixed-integer solutions at reduced runtime.

Load-bearing premise

The claim that moderate producer fairness improves business metrics rests on a purchase simulation where consumers buy each recommended item with probability equal to its relevance score and a producer is removed after one sale.

Editorial extensions

If this is right

Platforms showing multiple items per user must treat producer fairness as a real trade-off; the cost grows with list length k.
Moderate exposure guarantees can be a business lever: diversifying exposure away from saturated producers raises simulated sell-through rate and GMV.
Mean or max-min consumer objectives leave substantial inter-group variance; CVaR compresses it without meaningful utility loss elsewhere.
Relaxed LP with rounding and augmented Lagrangian match exact MIP quality at a fraction of the runtime, making fairness-aware re-ranking feasible at scale.

Reading between the lines

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

The business-improvement result is only as strong as the Appendix C purchase simulation (relevance-proportional Bernoulli buys, one-sale sell-out); a live marketplace test is the natural check.
An adaptive gamma tuned to market dynamics, rather than a manually fixed fairness level, is the natural next step the authors flag but do not solve.
The CVaR group objective could transfer to other allocation domains (ads placement, gig work, content moderation), where group definitions are contested and outcomes matter.
Near-perfect Precision@10 suggests the relevance signal is optimistic; with noisy production scores, the measured utility drops and fairness gains would likely shift.

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.

Desk Editor's Note

read the letter

Core message: this is a useful framework paper with a central observation that deserves attention, but the current empirical support is not as strong as the prose suggests.

The paper extends two-sided fairness from soft single-item allocation to discrete k>1, adds a CVaR group objective, and tests the 'free fairness' result from Greenwood et al. Across three datasets, it reports that at k=10 and full fairness, mean consumer utility drops 15–25%, while k=1 stays flat. That is a clean and important negative result if true — it says a well-known optimistic finding doesn't transfer to the setting platforms actually face. The CVaR extension is natural, and the runtime comparison of LP/rounding vs MIP is useful, though not surprising. Credit where due: code and data are promised on GitHub, and the limitations section is honest about offline simulation and positional bias.

Soft spots, in order of severity:

1. The GMV constraint in Eqs. 8 and 14 sums rho_ij, not w_ij. The constraint is independent of the allocation — it cannot bind on the optimization. This is not a typo; it is load-bearing for the business-constraint part of the paper. The GMV results in Figure 6 cannot be explained by the constraint as written.

2. Precision@10 of 97.4 and 97.7 with no train/test split almost certainly means the scores are evaluated on training interactions. If so, the relevance matrix is near-binary, and at k=10 the allocation experiments force many zero-relevance items into lists, mechanically creating the large utility drop. The central claim might still survive with realistic scores, but the magnitude and even the shape of the curves could change. The paper needs a proper held-out evaluation or a controlled noise injection.

3. The purchase simulation assumes consumers buy with probability equal to relevance and producers are removed after one sale. Diversifying exposure then trivially reduces sell-outs and raises STR/GMV. That is a modeling choice, but it cannot support the broad claim that fairness 'can improve business metrics' without a sensitivity analysis or a more realistic demand model.

4. Headline figures 3–6 have no error bars or significance tests. Table 1 has standard errors, but the main curves are single runs.

What is not a problem: the 'free fairness' claim is correctly attributed to prior work, and the CVaR formulation is properly credited to Williamson and Menon. The paper does not overclaim novelty.

Bottom line: the framework is a reasonable contribution, and the k>1 observation is an interesting hypothesis. But the evidence as presented is conditional. This deserves peer review because the question matters and the framework is reusable, but not acceptance in current form. The GMV constraint needs fixing, the recommender evaluation needs a clear train/test split, and the business simulation needs a robustness check. If the central result survives a proper held-out evaluation, it will be a solid paper.

Referee Report

5 major / 4 minor

Summary. The paper formalizes a two-sided recommendation allocation problem with discrete multi-item lists (w_ij ∈ {0,1}, k>1), producer exposure guarantees, group-level CVaR consumer fairness, and business (GMV) constraints. It proposes LP relaxation/rounding and gradient-based solvers, and evaluates on MovieLens, Amazon Reviews, and SimRec. The headline empirical claims are that the 'free fairness' regime of k=1 disappears at k=5/10 (15–25% utility drop at k=10, γ=1), moderate γ improves simulated STR/GMV, and CVaR compresses inter-group variance.

Significance. If the findings are robust, they would reframe fairness in two-sided markets as a potential business lever and caution against extrapolating single-item results. The paper ships code and includes realistic discrete allocations; these are strengths. However, the current empirical evidence is undermined by a GMV constraint that does not depend on the allocation, an inconsistent utility normalization, and a missing held-out evaluation of the relevance model, so the central claims are not yet established.

major comments (5)

[§4.3, Eq. (14) and Eq. (8)] The GMV constraint as printed is Σ_j v_j (Σ_i ρ_ij) ≥ θ V_max^GMV. This expression contains only ρ and v, not the allocation w. It therefore does not constrain w at all and cannot affect any optimization. Yet Section 5.1 reports GMV results under varying θ (Figure 6). The experiments must have used a different, unstated constraint (e.g., Σ_j v_j Σ_i w_ij), or the GMV findings are vacuous. This must be corrected and the experiments rerun.
[§5.1 / Appendix A, Table 3] No train/test split is reported for the two-tower recommender. Precision@10 of 97.41 (MovieLens) and 97.73 (Amazon) is far above typical held-out values, indicating likely evaluation on training interactions. If ρ is memorized, relevance scores are near-binary; at k>1, exposure constraints force zero-relevance items into lists and the 15–25% utility drop becomes a mechanical consequence, not a marketplace trade-off. The relevance matrices must be produced with a proper held-out split, and the paper should report standard test-set metrics.
[§4.1, Eq. (9)/(10) vs Eq. (1)] The paper's consumer utility definition normalizes by the top-k sum (Eq. 1), but the max-min and mean objectives in Eq. (9)–(10) normalize each term by max_j ρ_ij. For k>1 these denominators differ, so the utility values in Figure 3 and Table 1 are not computed according to the paper's own definition. This inconsistency affects every quantitative claim about utility costs and must be resolved.
[Appendix C / §5.1 (Fairness and Business Alignment)] The STR/GMV simulation assumes independent Bernoulli purchases with success probability ρ_ij and removes a producer after one sale. This contradicts Assumption A4 in §3.1 (unit supply, no depletion during allocation). It also mechanically produces higher sell-through with more diverse exposure: if a producer is removed after one sale, concentrating exposure causes early sell-outs by construction. To support the claim that fairness improves business metrics, the paper must test alternative purchase models (e.g., no removal, multi-unit inventory, session-level choice) and report sensitivity.
[Figures 3–6 and Tables 1–2] The central empirical curves are shown without error bars or significance tests. Table 1 reports ±0.00 for many cells despite only 3 runs. The claims that CVaR compresses variance and that fairness costs 15–25% need variability bounds and statistical tests across seeds, group partitions, and solver runs.

minor comments (4)

[§4.3, Eq. (13)] V_max^GMV is defined as the sum of the top k·n producer values, but the LHS of Eq. (14) contains an additional sum over consumers i of ρ_ij. The two sides are not dimensionally or semantically aligned; please clarify what quantity is being bounded.
[Figure 7] The y-axis label 'SCGard' should be 'SCGrad'.
[§3.1 (A4) vs Appendix C] Assumption A4 says items can be recommended to multiple consumers without depletion, but the purchase simulation removes a producer after one sale. Please state explicitly which setting applies to each part of the evaluation.
[Appendix A, Table 3] Reporting NDCG@10 and the number of test users would help readers judge the quality of the relevance model; Precision@10 alone is not standard for this claim.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the results are empirical outputs of constrained optimization and explicit simulations, not identities, fitted predictions, or self-citations.

full rationale

The paper's claim chain is empirical. It defines consumer utility, producer exposure, CVaR loss, and GMV in Section 3, then solves discrete allocation problems (Eqs. 9, 10, 12, 14) and measures the optimized outcomes. The 'free fairness disappears' result is not a definitional identity: it is the optimized value of the objective as the fairness parameter gamma increases, as shown in Figure 3. The CVaR result is likewise the value of the CVaR objective (12), not a tautology. Business-metric claims come from the explicit purchase simulation in Appendix C, which assumes Bernoulli purchase probability equal to relevance and one-sale producer removal; under that model, diversification reduces sell-outs, so improved STR is entailed by the model. This is a stated simulation assumption, not a fitted parameter disguised as a prediction, and the paper itself acknowledges the limitation: 'Our evaluation relies on offline simulations with fixed group structure.' No load-bearing author self-citation is present: prior work [12] is external and used as a baseline/contrast, not to justify the paper's constraints. The only validity concern is the near-perfect Precision@10 in Table 3 without a stated train/test split; if those scores were computed on training data, the magnitudes of the effects could be artifacts. However, that is an empirical validity threat, not a circular step: no equation in the paper reduces to its input by construction, and no fitted value is renamed as a prediction. Therefore the paper is not significantly circular; score 0.

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

The central claims rest on four explicit simplifications, a synthetic purchase simulation, and several hand-set optimization parameters. None of these are fitted to the headline outcomes, but the business-metric finding is particularly sensitive to the purchase model.

free parameters (7)

gamma (γ) = swept over [0,1]; moderate range γ≈0.3–0.6 highlighted
Producer fairness level; the business-improvement conclusion is drawn from the moderate part of this sweep.
alpha (α) = 0.95 in experiments
CVaR risk-aversion; higher values focus on worst groups.
theta (θ) = swept over [0,1]
GMV threshold in Eq. (8)/(14); controls the strength of the business constraint.
list length k = 1, 5, 10, 25
Recommendation list length; central to the claim that free fairness disappears for k>1.
group partition = MovieLens: 9 genre clusters; Amazon: 18 category clusters; SimRec: 10 Zipf groups
Hand-chosen partitions; the CVaR disparity results depend on these definitions.
SimRec generator parameters = Zipf exponent 1.0, noise σ=0.2, group-specific β_g, 1000x1000
Synthetic data generator used to probe controlled disparities; results may depend on these choices.
solver hyperparameters = λ, η0, anneal_rate, η_min (Appendix B)
AugLag/SCGrad tuning; not fully specified and affect runtime/violation results.

assumptions (7)

domain assumption A1: All slots in a recommendation list receive equal attention (no positional bias).
Stated in Section 3.1; isolates allocation from ranking effects, but breaks for ranked feeds as the authors acknowledge in the conclusion.
domain assumption A2: The probability a consumer engages with a recommended producer is proportional to the relevance score ρ_ij.
Stated in Section 3.1; the simulated purchase model in Appendix C uses this to generate STR/GMV results.
domain assumption A3: Each producer offers exactly one item.
Stated in Section 3.1; simplifies producer exposure to item exposure.
domain assumption A4: Unit supply: an item can be recommended to multiple consumers without depletion during allocation.
Stated in Section 3.1; appropriate for digital goods, but the STR simulation later removes sold-out producers, partially contradicting A4.
standard math Binary allocation with k per consumer and producer exposure constraints is NP-hard.
Invoked in Section 3.3/4 with citation [17] to justify approximate solvers.
domain assumption Relevance scores from the trained two-tower recommender are a meaningful input for downstream allocation.
Section 5 relies on ρ from MovieLens/Amazon; Table 3's near-perfect Precision@10 without a stated test split weakens this assumption.
domain assumption Maximum achievable GMV is the sum of the top k·n producer values (Eq. 13).
Normalization used in the GMV constraint; not justified beyond the k-slot counting.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Equity by Design? On the Trade-Offs in Fairness-Driven Recommendation in Heterogeneous Two-Sided Markets." pith.science (2026). https://pith.science/paper/BID4OK5Q

@misc{pith2026260210739,
  author       = {Pith},
  title        = {Pith review of: Equity by Design? On the Trade-Offs in Fairness-Driven Recommendation in Heterogeneous Two-Sided Markets},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BID4OK5Q}},
  note         = {Machine review of arXiv:2602.10739}
}

read the original abstract

Two-sided marketplaces embody heterogeneity in incentives: producers seek exposure while consumers seek relevance, and balancing these competing objectives through constrained optimization is now a standard practice. Yet practical platforms face interacting sources of heterogeneity that are often studied separately: multi-item recommendation, heterogeneous consumer groups, and business constraints beyond raw relevance. In this work, we present and study offline optimization framework for analyzing these trade-offs in an unified manner, extending prior two-sided formulations to represent more realistic discrete multi-item recommendations. Within this framework, we couple producer-side exposure guarantees with a consumer-group fairness objective and explicit business-oriented constraints. Our experiments show that the previously reported ``free fairness'' regime from highly stylized single-item recommendation settings disappears once each consumer receives multiple recommendations, and that moderate producer-fairness constraints can improve simulated business metrics by diversifying exposure away from saturated producers. We further show that reduction of inter-group disparity, preserves competitive overall utility.

Figures

Figures reproduced from arXiv: 2602.10739 by the authors.

**Figure 3.** Mean consumer utility versus producer fairness [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

**Figure 4.** Group-level consumer utilities under max-min, [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗

Figures from the paper (4 more)

**Figure 7.** Figure 7: Constraint violations for AugLag and SCGrad as a [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]

**Figure 6.** Figure 6: Simulated GMV as a function of business-value con [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]

**Figure 8.** Figure 8: Average producer relevance decay across consumer [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]

**Figure 9.** Figure 9: Producer allocation patterns on Amazon Reviews [PITH_FULL_IMAGE:figures/full_fig_p009_9.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

36 extracted references · 3 canonical work pages

[1]

Himan Abdollahpouri, Gediminas Adomavicius, Robin Burke, Ido Guy, Dietmar Jannach, Toshihiro Kamishima, Jan Krasnodebski, and Luiz Pizzato. 2020. Mul- tistakeholder recommendation: Survey and research directions.User Modeling and User-Adapted Interaction30 (2020), 127–158

2020
[2]

Philippe Artzner, Freddy Delbaen, Jean-Marc Eber, and David Heath. 1999. Co- herent measures of risk.Mathematical finance9, 3 (1999), 203–228

1999
[3]

Theo Barre, Omar El Housni, Marouane Ibn Brahim, Andrea Lodi, and Danny Segev. 2023. Assortment Optimization with Visibility Constraints. arXiv:arXiv:2307.13656

arXiv 2023
[4]

Kinjal Basu, Cyrus DiCiccio, Heloise Logan, and Noureddine El Karoui. 2020. A Framework for Fairness in Two-Sided Marketplaces. arXiv:2006.12756 [cs] doi:10.48550/arXiv.2006.12756

work page Pith review arXiv doi:10.48550/arxiv.2006.12756 2020
[5]

Patro, Niloy Ganguly, Krishna P

Arpita Biswas, Gourab K. Patro, Niloy Ganguly, Krishna P. Gummadi, and Ab- hijnan Chakraborty. 2022. Towards Fair Recommendation in Two-Sided Plat- forms.ACM Transactions on the Web16, 2 (May 2022), 1–34. doi:10.1145/3503624 arXiv:2201.01180 [cs]

arXiv 2022
[6]

Jesús Bobadilla, Fernando Ortega, Antonio Hernando, and Abraham Gutiérrez
[7]

Suresh Bolusani, Mathieu Besançon, Ksenia Bestuzheva, Antonia Chmiela, João Dionísio, Tim Donkiewicz, Jasper van Doornmalen, Leon Eifler, Mohammed Ghannam, Ambros Gleixner, et al. 2024. The SCIP optimization suite 9.0.arXiv preprint arXiv:2402.17702(2024)

arXiv 2024
[8]

Javier Cembrano, Andrés Moraga, and Victor Verdugo. 2025. Near-feasible Fair Allocations in Two-sided Markets. arXiv:arXiv:2506.01178

arXiv 2025

Show all 36 references

[9]

Qinyi Chen, Jason Cheuk Nam Liang, Negin Golrezaei, and Djallel Bouneffouf
[10]

Paul Covington, Jay Adams, and Emre Sargin. 2016. Deep Neural Networks for YouTube Recommendations. InProceedings of the 10th ACM Conference on Recommender Systems(Boston, Massachusetts, USA)(RecSys ’16). Association for Computing Machinery, New York, NY, USA, 191–198. doi:10....

2016 doi
[11]

Yashar Deldjoo, Dietmar Jannach, Alejandro Bellogin, Alessandro Difonzo, and Dario Zanzonelli. 2024. Fairness in recommender systems: research landscape and future directions.User Modeling and User-Adapted Interaction34, 1 (2024), 59–108

2024
[12]

Sophie Greenwood, Sudalakshmee Chiniah, and Nikhil Garg. 2024. User-item fairness tradeoffs in recommendations.Advances in Neural Information Processing Systems(2024)

2024
[13]

Gurobi Optimization, LLC. 2024. Gurobi Optimizer Reference Manual. https: //www.gurobi.com

2024
[14]

F Maxwell Harper and Joseph A Konstan. 2015. The movielens datasets: History and context.Acm transactions on interactive intelligent systems (tiis)5, 4 (2015), 1–19

2015
[15]

Yupeng Hou, Jiacheng Li, Zhankui He, An Yan, Xiusi Chen, and Julian McAuley
[16]

Aryan Jadon and Avinash Patil. 2024. A comprehensive survey of evaluation tech- niques for recommendation systems. InInternational Conference on Computation of Artificial Intelligence & Machine Learning. Springer, 281–304

2024
[17]

2011.Combinatorial optimization

Bernhard H Korte, Jens Vygen, B Korte, and J Vygen. 2011.Combinatorial optimization. Vol. 1. Springer

2011
[18]

Bridging language and items for retrieval and recommendation.arXiv preprint arXiv:2403.03952(2024)

2024 arXiv
[19]

Yang Li, Kangbo Liu, Ranjan Satapathy, Suhang Wang, and Erik Cambria. 2023. Recent developments in recommender systems: A survey (2023).arXiv preprint arXiv:2306.12680(2023)

2023 arXiv
[20]

Xinyu Lin, Wenjie Wang, Jujia Zhao, Yongqi Li, Fuli Feng, and Tat-Seng Chua
[21]

Yunqi Li, Hanxiong Chen, Shuyuan Xu, Yingqiang Ge, Juntao Tan, Shuchang Liu, and Yongfeng Zhang. 2023. Fairness in recommendation: Foundations, methods, and applications.ACM Transactions on Intelligent Systems and Technology14, 5 (2023), 1–48

2023
[22]

Yuanzhe Peng. 2022. A survey on modern recommendation system based on big data.arXiv preprint arXiv:2206.02631(2022). Seputis et al

2022 arXiv
[23]

Marco Tulio Ribeiro, Nivio Ziviani, Edleno Silva De Moura, Itamar Hata, Anisio Lacerda, and Adriano Veloso. 2015. Multiobjective Pareto-Efficient Approaches for Recommender Systems.ACM Transactions on Intelligent Systems and Tech- nology5, 4 (Jan. 2015), 1–20. doi:10.1145/2629350

2015 doi
[24]

arXiv:2312.09901 [cs] doi:10.48550/arXiv.2312.09901

Temporally and Distributionally Robust Optimization for Cold-Start Rec- ommendation. arXiv:2312.09901 [cs] doi:10.48550/arXiv.2312.09901

doi arXiv
[25]

Phong Nguyen, John Dines, and Jan Krasnodebski. 2017. A Multi-Objective Learning to Re-Rank Approach to Optimize Online Marketplaces for Multiple Stakeholders. arXiv:1708.00651 [cs] doi:10.48550/arXiv.1708.00651

2017 doi arXiv
[26]

Yifan Wang, Weizhi Ma, Min Zhang, Yiqun Liu, and Shaoping Ma. 2023. A survey on the fairness of recommender systems.ACM Transactions on Information Systems41, 3 (2023), 1–43

2023
[27]

Yifan Wang, Peijie Sun, Weizhi Ma, Min Zhang, Yuan Zhang, Peng Jiang, and Shaoping Ma. 2024. Intersectional Two-sided Fairness in Recommendation. (2024). arXiv:arXiv:2402.02816 doi:10.1145/3589334.3645518

2024 arXiv
[28]

R Tyrrell Rockafellar, Stanislav Uryasev, et al. 2000. Optimization of conditional value-at-risk.Journal of risk2 (2000), 21–42

2000
[29]

Ashudeep Singh and Thorsten Joachims. 2018. Fairness of exposure in rankings. InProceedings of the 24th ACM SIGKDD international conference on knowledge discovery & data mining. 2219–2228

2018
[30]

Yuying Zhao, Yu Wang, Yunchao Liu, Xueqi Cheng, Charu C Aggarwal, and Tyler Derr. 2023. Fairness and diversity in recommender systems: a survey.ACM Transactions on Intelligent Systems and Technology(2023)

2023
[31]

Yong Zheng. 2017. Multi-stakeholder recommendation: Applications and chal- lenges.arXiv preprint arXiv:1707.08913(2017)

2017 arXiv
[32]

Williamson and Aditya Krishna Menon

Robert C. Williamson and Aditya Krishna Menon. 2019. Fairness Risk Measures. arXiv:1901.08665 [cs] doi:10.48550/arXiv.1901.08665

2019 doi arXiv
[33]

Zikun Ye, Reza Yousefi Maragheh, Lalitesh Morishetti, Shanu Vashishtha, Jason Cho, Kaushiki Nag, Sushant Kumar, and Kannan Achan. 2023. Seller-side Out- come Fairness in Online Marketplaces.arXiv preprint arXiv:2312.03253(2023)

2023 arXiv
[36]

Quan Zhou, Jakub Marecek, and Robert N. Shorten. 2021. Subgroup Fair- ness in Two-Sided Markets. PLoS ONE 18(2): e0281443, 2023. (2021). arXiv:arXiv:2106.02702 doi:10.1371/journal.pone.0281443

2021 arXiv
[2013]

Recommender systems survey.Knowledge-based systems46 (2013), 109– 132

2013
[2024]

Advances in Neural Information Processing Systems37 (2024), 50189–50229

Interpolating item and user fairness in multi-sided recommendations. Advances in Neural Information Processing Systems37 (2024), 50189–50229

2024

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.

[1] [1]

Himan Abdollahpouri, Gediminas Adomavicius, Robin Burke, Ido Guy, Dietmar Jannach, Toshihiro Kamishima, Jan Krasnodebski, and Luiz Pizzato. 2020. Mul- tistakeholder recommendation: Survey and research directions.User Modeling and User-Adapted Interaction30 (2020), 127–158

2020

[2] [2]

Philippe Artzner, Freddy Delbaen, Jean-Marc Eber, and David Heath. 1999. Co- herent measures of risk.Mathematical finance9, 3 (1999), 203–228

1999

[3] [3]

Theo Barre, Omar El Housni, Marouane Ibn Brahim, Andrea Lodi, and Danny Segev. 2023. Assortment Optimization with Visibility Constraints. arXiv:arXiv:2307.13656

arXiv 2023

[4] [4]

Kinjal Basu, Cyrus DiCiccio, Heloise Logan, and Noureddine El Karoui. 2020. A Framework for Fairness in Two-Sided Marketplaces. arXiv:2006.12756 [cs] doi:10.48550/arXiv.2006.12756

work page Pith review arXiv doi:10.48550/arxiv.2006.12756 2020

[5] [5]

Patro, Niloy Ganguly, Krishna P

Arpita Biswas, Gourab K. Patro, Niloy Ganguly, Krishna P. Gummadi, and Ab- hijnan Chakraborty. 2022. Towards Fair Recommendation in Two-Sided Plat- forms.ACM Transactions on the Web16, 2 (May 2022), 1–34. doi:10.1145/3503624 arXiv:2201.01180 [cs]

arXiv 2022

[6] [6]

Jesús Bobadilla, Fernando Ortega, Antonio Hernando, and Abraham Gutiérrez

[7] [7]

Suresh Bolusani, Mathieu Besançon, Ksenia Bestuzheva, Antonia Chmiela, João Dionísio, Tim Donkiewicz, Jasper van Doornmalen, Leon Eifler, Mohammed Ghannam, Ambros Gleixner, et al. 2024. The SCIP optimization suite 9.0.arXiv preprint arXiv:2402.17702(2024)

arXiv 2024

[8] [8]

Javier Cembrano, Andrés Moraga, and Victor Verdugo. 2025. Near-feasible Fair Allocations in Two-sided Markets. arXiv:arXiv:2506.01178

arXiv 2025