Pith. sign in

REVIEW 5 major objections 5 minor 50 references

Plausible Counterfactual Explanations of Recommendations

T0 review · 5 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read Recommender counterfactuals can be found in seconds and are judged plausible by users.

desk verdict Solid MIO formulation and public code, but the plausibility claim is undercut by a self-referential SPN metric and a user study that omits the optimize variant. read the letter →

arxiv 2507.07919 v1 pith:WKLQ7SX3 submitted 2025-07-10 cs.LG cs.IR

Jakub Černý , Jiří Němeček , Ivan Dovica , Jakub Mareček This is my paper · ORCID
classification cs.LGcs.IR
keywords counterfactualexplanationsrecommendersystemssum-productnetworksmixed-integeroptimizationplausibilitytop-krecommendationexplainabilityuserstudy
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 often explain a suggestion by counterfactual: "you were recommended X because you interacted with Y; if you had not, X would drop off the list." This paper argues that such explanations should not only be close to the user's current behavior but also plausible, meaning likely under the data distribution. It adapts the LiCE method to recommenders by encoding the top-k ranking condition as mixed-integer optimization and using a sum-product network to estimate counterfactual likelihood, testing several ways of aggregating items into categories. On binarized interactions with disjunction-aggregated features, it reliably finds counterfactuals that drop the item from the top-k list, with a median solve time of 11 seconds, and a 20-person user study found the resulting explanations sensible and helpful for understanding the recommender.

What carries the argument

The machinery is a mixed-integer optimization model whose objective is the $\ell^1$ distance from the user's current interaction vector, subject to constraints that encode the recommender's scoring function and a rank-based "drop out of top-k" condition built with binary ranking variables and a big-M constant. A sum-product network (SPN), optionally trained on category-aggregated features using sum, mean, or disjunction aggregation, supplies the likelihood term that makes counterfactuals plausible. The rank constraints ensure the explanation actually removes the item from the recommendation list, while the SPN term pushes the search toward changes that look like the training data.

What would settle it

Collect paired counterfactuals for the same user and item that have identical category-aggregated SPN likelihood but differ in which specific item within a category is removed (for example, removing a frequently rated versus a rarely rated merchant in the same category); if users do not judge the two as equally plausible, the category-level likelihood is not measuring item-level plausibility and the method's central promise fails.

Watch

Extended reading notes

Core claim

The author's central claim is that counterfactual explanation generation for top-k recommenders can be formulated as a mixed-integer optimization problem, with a sum-product network scoring the likelihood of candidate counterfactuals, and that this reliably produces high-likelihood, model-faithful counterfactuals without exhaustive search. The rank-based constraint guarantees the explained item actually leaves the top-k list, which a score-threshold constraint alone does not ensure. The paper demonstrates this with EASE as the recommender across three public datasets and a private transaction dataset, and reports that users rate the resulting explanations as sensible.

Load-bearing premise

The SPN trained on category-aggregated feature vectors correctly estimates how plausible a change to the original item-level interactions looks, even though the paper does not independently verify that the two agree.

Editorial extensions

If this is right

  • For EASE-style linear recommenders, a counterfactual that drops an item from the top-k list can be computed by an open-source solver in a median of 11 seconds, making the approach usable when explanations are generated on request.
  • Guaranteeing the drop by a rank constraint matters: with only a score-threshold constraint, roughly 30% of the generated counterfactuals still leave the item inside the top-k.
  • On binarized interaction data, aggregating item features by disjunction makes the optimization easiest to solve and the learned SPN least restrictive, while rating-based data with mean aggregation often makes the problem infeasible.
  • Using the SPN likelihood as an objective term (with weight $\alpha = 0.1$) avoids the frequent infeasibility of the hard median-likelihood threshold variant.
  • In the user study, respondents rated all tested explanation setups above the midpoint of the sensible scale, and the majority agreed that the explanations made the recommender's decisions more transparent and understandable.

Reading between the lines

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

  • If category-level likelihood does not track item-level plausibility, the method's plausibility guarantees are weaker than they appear; a direct test would be to hold the aggregated SPN likelihood fixed while swapping a frequently-rated item for a rarely-rated one in the same category and asking users which counterfactual is more sensible.
  • The user study's preference for the fastest, sparsest setup without an SPN hints that users may weigh conciseness over distributional likelihood; an approach that minimizes $\ell^1$ distance first and then re-ranks the few best candidates by SPN likelihood could capture both.
  • The same MIO-plus-SPN recipe could apply to other decision problems with discrete inputs and tractable scoring functions, such as loan approval, admission, or hiring, where "what would change the decision?" questions have the same top-k or threshold structure.
  • Because the method only allows interaction values to decrease, it is tailored to "you were recommended X because of Y" explanations; allowing increases would require additional monotonicity assumptions in the recommender score.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 5 minor

Summary. The paper presents an extension of the LiCE method for generating counterfactual explanations (CEs) in recommender systems. The authors formulate CE generation as a mixed-integer optimization problem with a k-th rank constraint (Eqs. 5-6), use sum-product networks (SPNs) to model plausibility, and propose three category-level aggregation functions (sum, mean, disjunction) to make SPN training tractable. They compare three ways of using the SPN (optimize likelihood with weight alpha=0.1, constrain likelihood to be above a threshold, and no SPN) across three public datasets and one private dataset, report run-time and plausibility/similarity trade-offs (Tables I-II), and supplement the numerical evaluation with a user study on merchant recommendations (Section VI). The central claims are that the method reliably finds close, faithful counterfactuals without exhaustive search and that the generated CEs have high likelihood with respect to the data distribution.

Significance. If the main claims held, the paper would make a useful contribution: it provides an openly available, MIO-based CE generator tailored to recommender rankings, introduces a k-th rank validity constraint that is more faithful to top-k recommendation than simple score thresholds, and reports a user study with statistical tests. The open-source implementation and the explicit treatment of the ranking constraint are commendable. However, the central plausibility claim currently rests on a self-referential evaluation: the SPN used to measure plausibility is the same category-aggregated SPN that the Optimize variant directly optimizes, and the user study does not test that Optimize variant. These gaps are load-bearing for the paper's main contribution and require additional experiments or careful re-scoping before the claims can be accepted.

major comments (5)
  1. [§IV-A, §V-A, Table II] The plausibility metric in Table II is the log-likelihood under the same category-aggregated SPN that the Optimize variant is explicitly trained to maximize in the objective. Therefore, the reported improvement in plausibility for the SPN variants is partly a consequence of optimizing the evaluation metric itself, not an independent measurement of plausibility. The manuscript should validate the SPN externally, for example with a held-out SPN, an item-level density model, or human ratings that include the Optimize variant.
  2. [§IV-A, Eqs. (7)-(11)] The SPN is defined on category-aggregated vectors z_i, while the counterfactual changes the item-level vector x_i. The aggregation is many-to-one: with the disjunction aggregator, replacing any item in a category by another item in the same category leaves z_i unchanged, and with the sum aggregator, exchanging a popular item for an obscure item within a category is invisible to the SPN. Consequently, high category-level likelihood does not imply that the item-level counterfactual is plausible. The paper should provide evidence that p(z) tracks p(x), for instance by comparing category-level likelihoods with item-level likelihoods on a validation set.
  3. [§VI, Table IV and Fig. 2] The user study does not include the Optimize variant (the alpha=0.1 likelihood term in the objective), which is the main numerical contribution of the paper. The best user-rated approach, C, uses no SPN at all, and the best SPN-based rank-guarantee approach, A, is statistically equivalent to C rather than superior. Thus the user study does not provide external support for the claim that SPN-based optimization improves perceived plausibility; the authors should either rerun the study with the Optimize variant or clearly limit their claims to the configurations actually tested.
  4. [§V, Tables I-II] All numerical experiments use EASE as the recommender, although Section III-b claims the approach improves on the state of the art for 'most recommender models' and Section IV discusses formulations for NCF and VASP. The general claim is not supported by experiments with a single recommender model; testing at least one additional model, such as NCF or a VAE-based recommender, is needed to substantiate the claimed generality.
  5. [§V-A, Tables I-II] The paper selects the disjunction aggregator after inspecting Table I and then reports only that configuration in Table II. This post-hoc selection, made without a separate validation split, means the reported plausibility and similarity comparisons are not a clean test of the method as specified. The authors should report results for all aggregators or justify the selection on a priori grounds.
minor comments (5)
  1. [§IV, Eq. (5)] The text around Eq. (5) says 'the merchant j ranks above the merchant i,' but i already denotes the user; the explained item should be c, not i.
  2. [§IV] The sentence 'In the continuous case, one can introduce [13], [14] auxiliary variable' is missing an article; it should read 'an auxiliary variable.'
  3. [§VI, Table IV] There are typos in Table IV and the surrounding text: '5 positions dorp' should be '5 positions drop,' and 'recommendatoin' should be 'recommendation.'
  4. [§VI] The text contains the typo 'fied of business' for 'field of business.'
  5. [Table II] Several numeric entries in Table II have stray spacing (e.g., '1 .16±0.49'); the formatting should be made consistent.

Circularity Check

2 steps flagged · score 6.0 of 10

Plausibility is measured by the same SPN log-likelihood that the Optimize variant minimizes, making the reported plausibility advantage partly tautological and unvalidated at the item level.

  1. fitted input called prediction [Section V (Numerical Evaluation), objective (1) and Table II]
    "we either impose a constraint on the CE likelihood to be at least equal to the median likelihood of the training data, or we subtract the likelihood into the objective (1) multiplied by α = 0.1, effectively minimizing a linear combination of ℓ1 distance and negative log-likelihood."

    The Optimize variant's objective is a linear combination of ℓ1 distance and negative log-likelihood under the SPN. Table II reports 'Plausibility (Log-Likelihood)' as the evaluation metric, which is the negation of that same objective term. Therefore, the improvement in plausibility of Optimize over No SPN is a direct consequence of optimizing the evaluation metric itself: the variant was constructed to minimize -log p(z), and then evaluated on log p(z). The Threshold variant is similarly forced to have likelihood at least the training-data median, so its high plausibility scores are also built into the constraint. No independent item-level plausibility metric or human evaluation of the Optimize variant is provided.

  2. self definitional [Section I (Introduction), Section IV-A, Section V (Numerical Evaluation)]
    "we define plausibility as the likelihood of the counterfactual with respect to the training data. ... We can then aggregate the sparse information into K-dimensional feature vectors and use those for training. ... To model the data distribution, working with the full input space would be inefficient. We therefore group items into the categories provided with the data."

    Plausibility is defined with respect to the training data, i.e., the item-level interaction vectors x_i, but the SPN is trained on category-aggregated vectors z_i (sum, mean, or disjunction, Eqs. 7-11). The mapping from x to z is many-to-one: for disjunction, changing any one item in a category yields the same z, and for sum, replacing a popular item by an obscure item within a category is invisible. The numerical evaluation in Table II uses log-likelihood under this same aggregated SPN, so the paper's central claim of 'high likelihood with respect to the data distribution' is self-referential: it never checks whether p(z) tracks p(x). The claim of item-level plausibility is thus supported only by a proxy that is not validated independently.

full rationale

The paper proposes a modified LiCE for recommender systems, using an SPN over either item-level or category-aggregated features to enforce plausibility, and compares Optimize, Threshold, and No-SPN variants. The central numerical claim—that the SPN variants improve plausibility—rests on Table II's 'Plausibility (Log-Likelihood)' metric, which is the same SPN log-likelihood that the Optimize variant minimizes (via α=0.1 negative log-likelihood in the objective) and that the Threshold variant constrains. Hence the improvement is partly by construction rather than an independent finding. Furthermore, the SPN is trained on aggregated category features (Eqs. 7-11), and the evaluation uses that same aggregated likelihood without any validation that category-level p(z) reflects item-level p(x), so the definition of 'data distribution' is silently shifted from item-level training data to aggregated vectors. The user study is independent human feedback, but it does not rescue the SPN-specific claim: approach C (no SPN) is rated significantly better than B, D, and E, and statistically equivalent to A, so the non-circular evidence actually undermines the claim that SPN variants produce the most plausible explanations. The MIO formulation itself and the validity guarantees are not circular, but the plausibility claims reduce substantially to the metric being optimized. Score 6 reflects that the numerical plausibility result is partially forced by the evaluation setup, while the underlying optimization framework retains independent content.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

No new entities are introduced. The free parameters are the objective weight alpha and the likelihood threshold. The key assumptions are MIO representability of scoring functions, decrease-only counterfactuals, category-level SPN as likelihood proxy, and standard MIO techniques.

free parameters (2)
  • alpha = 0.1
    Weight on the negative log-likelihood term in the optimize objective. Chosen by hand, not tuned or justified.
  • likelihood threshold = median log-likelihood of training data
    Used in the threshold variant; derived from data rather than a fixed constant, effectively a free choice of operating point.
assumptions (4)
  • domain assumption The recommender score function fj(x) can be represented exactly in mixed-integer optimization.
    Stated in Section IV; verified only for EASE in the experiments, so generalizability to NCF, VAE, and VASP is assumed.
  • domain assumption Counterfactuals are restricted to decreasing interaction values.
    Section V states only decreases are allowed to produce explanations of the form 'recommended X because you interacted with Y'. This excludes plausible counterfactuals that add interactions.
  • domain assumption The SPN trained on category-aggregated features approximates the likelihood of original item vectors.
    Section IV-A aggregates items into category sums, means, or disjunctions; the paper does not validate that this preserves the true data likelihood.
  • standard math Mixed-integer programming linearization and big-M constraints are valid.
    Standard MIO techniques from [14] used in Equations (5)-(6); assumed correct.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Plausible Counterfactual Explanations of Recommendations." pith.science (2026). https://pith.science/paper/WKLQ7SX3

@misc{pith2026250707919,
  author       = {Pith},
  title        = {Pith review of: Plausible Counterfactual Explanations of Recommendations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WKLQ7SX3}},
  note         = {Machine review of arXiv:2507.07919}
}
read the original abstract

Explanations play a variety of roles in various recommender systems, from a legally mandated afterthought, through an integral element of user experience, to a key to persuasiveness. A natural and useful form of an explanation is the Counterfactual Explanation (CE). We present a method for generating highly plausible CEs in recommender systems and evaluate it both numerically and with a user study.

Figures

Figures reproduced from arXiv: 2507.07919 by the authors.

Figure 1
Figure 1. Example of a counterfactual explanation relevance is estimated based on the user’s transaction history, i.e., looking at which merchants were visited (and how often) by each customer. The qualitative evaluation of the provided CEs is non-trivial. Transactions with suggested merchant and CE merchants may be related to various phenomena, such as the fied of business of the merchants, spatial patterns, or other user-sp… view at source ↗
Figure 3
Figure 3. Q: After seeing the explanations, I better understand the merchant [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 2
Figure 2. Evaluation of CE explanations in terms of how sensible are they to [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

50 extracted references · 27 canonical work pages

  1. [1]

    Explaining collaborative filtering recommendations,

    J. L. Herlocker, J. A. Konstan, and J. Riedl, “Explaining collaborative filtering recommendations,” in Proceedings of the 2000 ACM conference on Computer supported cooperative work . Philadelphia Pennsylvania USA: ACM, Dec. 2000, pp. 241–250. [Online]. Available: https://dl.acm.org/doi/10.1145/358916.358995

    work page arXiv

  2. [2]

    Why Am I Not Seeing It? Understanding Users’ Needs for Counterfactual Explanations in Everyday Recommendations,

    R. Shang, K. J. K. Feng, and C. Shah, “Why Am I Not Seeing It? Understanding Users’ Needs for Counterfactual Explanations in Everyday Recommendations,” in 2022 ACM Conference on Fairness, Accountability, and Transparency . Seoul Republic of Korea: ACM, Jun. 2022, pp. 1330–1340. [Online]. Available: https://dl.acm.org/doi/10.1145/3531146.3533189

    arXiv 2022

  3. [3]

    A systematic review and taxonomy of explanations in decision support and recommender systems,

    I. Nunes and D. Jannach, “A systematic review and taxonomy of explanations in decision support and recommender systems,” User Modeling and User-Adapted Interaction , vol. 27, no. 3, pp. 393–444, Dec. 2017. [Online]. Available: https://doi.org/10.1007/ s11257-017-9195-0

    work page 2017

  4. [4]

    Explore, exploit, and explain: personalizing explainable recommendations with bandits,

    J. McInerney, B. Lacker, S. Hansen, K. Higley, H. Bouchard, A. Gruson, and R. Mehrotra, “Explore, exploit, and explain: personalizing explainable recommendations with bandits,” in Proceedings of the 12th ACM Conference on Recommender Systems , ser. RecSys ’18. New York, NY , USA: Association for Computing Machinery, Sep. 2018, pp. 31–39. [Online]. Availab...

    arXiv 2018

  5. [5]

    Regulation (EU) 2016/679, “Regulation (EU) 2016/679 of the European Parliament and of the Council. of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC (General Data Protection Regulation),” 2016, place: OJ L 119, 4.5.2016, p. 1–88

    work page 2016

  6. [6]

    California consumer privacy act (ccpa),

    “California consumer privacy act (ccpa),” Retrieved from State of California Department of Justice , 2022. [Online]. Available: https://oag.ca.gov/privacy/ccpa

    work page 2022

  7. [7]

    Regulation (EU) 2024/1689, “Regulation (EU) 2024/1689 of the Euro- pean Parliament and of the Council. of 13 June 2024 laying down har- monised rules on artificial intelligence and amending Regulations (EC) No 300/2008, (EU) No 167/2013, (EU) No 168/2013, (EU) 2018/858, (EU) 2018/1139 and (EU) 2019/2144 and Directives 2014/90/EU, (EU) 2016/797 and (EU) 20...

    work page 2024

  8. [8]

    A Survey on Trustworthy Recommender Systems,

    Y . Ge, S. Liu, Z. Fu, J. Tan, Z. Li, S. Xu, Y . Li, Y . Xian, and Y . Zhang, “A Survey on Trustworthy Recommender Systems,” ACM Transactions on Recommender Systems , vol. 3, no. 2, pp. 1–68, Jun. 2024. [Online]. Available: https://dl.acm.org/doi/10.1145/3652891

    doi:10.1145/3652891 2024

Show all 50 references
  1. [9]

    Counterfactual Explanations Without Opening the Black Box: Automated Decisions and the GDPR,

    S. Wachter, B. Mittelstadt, and C. Russell, “Counterfactual Explanations Without Opening the Black Box: Automated Decisions and the GDPR,” SSRN Electronic Journal , 2017. [Online]. Available: https: //www.ssrn.com/abstract=3063289

    2017

  2. [10]

    CK v Dun & Bradstreet Austria GmbH and Magistrat der Stadt Wien,

    EU Court of Justice, “CK v Dun & Bradstreet Austria GmbH and Magistrat der Stadt Wien,” Feb. 2025

    2025

  3. [11]

    Counterfactual explanations and how to find them: literature review and benchmarking,

    R. Guidotti, “Counterfactual explanations and how to find them: literature review and benchmarking,” Data Mining and Knowledge Discovery, Apr. 2022. [Online]. Available: https://doi.org/10.1007/ s10618-022-00831-6

    2022

  4. [13]

    Generating likely counterfactuals using sum-product networks,

    J. N ˇemeˇcek, T. Pevn ´y, and J. Mare ˇcek, “Generating likely counterfactuals using sum-product networks,” in The thirteenth international conference on learning representations , 2025. [Online]. Available: https://openreview.net/forum?id=rGyi8NNqB0

    2025

  5. [14]

    L. A. Wolsey, Integer programming, second edition ed. Hoboken, NJ: Wiley, 2021

    2021

  6. [15]

    Progress in math- ematical programming solvers from 2001 to 2020,

    T. Koch, T. Berthold, J. Pedersen, and C. Vanaret, “Progress in math- ematical programming solvers from 2001 to 2020,” EURO Journal on Computational Optimization, vol. 10, p. 100031, Jan. 2022

    2001

  7. [16]

    Sum-product networks: A new deep architecture,

    H. Poon and P. Domingos, “Sum-product networks: A new deep architecture,” in 2011 IEEE International Conference on Computer Vision Workshops (ICCV Workshops), Nov. 2011, pp. 689–690. [Online]. Available: https://ieeexplore.ieee.org/abstract/document/6130310

    2011

  8. [17]

    Characteristic circuits,

    Z. Yu, M. Trapp, and K. Kersting, “Characteristic circuits,” in Advances in Neural Information Processing Systems , A. Oh, T. Naumann, A. Globerson, K. Saenko, M. Hardt, and S. Levine, Eds., vol. 36. Curran Associates, Inc., 2023, pp. 34 074–34 086. [Online]. Available: https:...

    2023

  9. [18]

    Probabilistic flow circuits: Towards unified deep models for tractable probabilistic inference,

    S. Sidheekh, K. Kersting, and S. Natarajan, “Probabilistic flow circuits: Towards unified deep models for tractable probabilistic inference,” in Uncertainty in Artificial Intelligence . PMLR, 2023, pp. 1964–1973

    2023

  10. [19]

    Explanation in Recommender Systems,

    D. Mcsherry, “Explanation in Recommender Systems,” Artificial Intelligence Review, vol. 24, no. 2, pp. 179–197, Oct. 2005. [Online]. Available: https://doi.org/10.1007/s10462-005-4612-x

    2005 doi

  11. [20]

    A Taxonomy for Gen- erating Explanations in Recommender Systems,

    G. Friedrich and M. Zanker, “A Taxonomy for Gen- erating Explanations in Recommender Systems,” AI Mag- azine, vol. 32, no. 3, pp. 90–98, 2011, eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1609/aimag.v32i3.2365. [On- line]. Available: https://onlinelibrary.wiley.com/doi/a...

    2011 doi

  12. [21]

    A survey on effects of adding explanations to recommender systems,

    A. Vultureanu-Albis ¸i and C. B ˘adic˘a, “A survey on effects of adding explanations to recommender systems,” Concurrency and Computation: Practice and Experience , vol. 34, no. 20, p. e6834, 2022, eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/cpe.6834. [Online]. Ava...

    2022 doi

  13. [22]

    Evaluating the Pros and Cons of Recommender Systems Explanations,

    K. Wardatzky, “Evaluating the Pros and Cons of Recommender Systems Explanations,” in Proceedings of the 18th ACM Conference on Recommender Systems , ser. RecSys ’24. New York, NY , USA: Association for Computing Machinery, Oct. 2024, pp. 1302–1307. [Online]. Available: https:/...

    2024

  14. [23]

    PRINCE: Provider-side Interpretability with Counterfactual Explanations in Recommender Systems,

    A. Ghazimatin, O. Balalau, R. Saha Roy, and G. Weikum, “PRINCE: Provider-side Interpretability with Counterfactual Explanations in Recommender Systems,” in Proceedings of the 13th International Conference on Web Search and Data Mining . Houston TX USA: ACM, Jan. 2020, pp. 196–...

    2020

  15. [24]

    Counterfactual Explanations for Neural Recommenders,

    K. H. Tran, A. Ghazimatin, and R. Saha Roy, “Counterfactual Explanations for Neural Recommenders,” in Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval . Virtual Event Canada: ACM, Jul. 2021, pp. 1627–1631. [Online]...

    2021

  16. [25]

    Learning Causal Explanations for Recommendation,

    S. Xu, Y . Li, S. Liu, Z. Fu, Y . Ge, X. Chen, and Y . Zhang, “Learning Causal Explanations for Recommendation,” 2021

    2021

  17. [26]

    Prototype-Guided Counterfactual Explanations via Variational Auto-encoder for Recommendation,

    M. He, J. Wang, B. An, and H. Wen, “Prototype-Guided Counterfactual Explanations via Variational Auto-encoder for Recommendation,” in Machine Learning and Knowledge Discovery in Databases: Applied Data Science and Demo Track , G. De Francisci Morales, C. Perlich, N. Ruchansky,...

    2023

  18. [27]

    Counterfactual Explainable Recommendation,

    J. Tan, S. Xu, Y . Ge, Y . Li, X. Chen, and Y . Zhang, “Counterfactual Explainable Recommendation,” in Proceedings of the 30th ACM International Conference on Information & Knowledge Management . Virtual Event Queensland Australia: ACM, Oct. 2021, pp. 1784–1793. [Online]. Avai...

    2021

  19. [28]

    Learning and evaluating graph neural network explanations based on counterfactual and factual reasoning,

    J. Tan, S. Geng, Z. Fu, Y . Ge, S. Xu, Y . Li, and Y . Zhang, “Learning and evaluating graph neural network explanations based on counterfactual and factual reasoning,” in Proceedings of the ACM web conference 2022 , ser. Www ’22. ACM, Apr. 2022, pp. 1018–1027. [Online]. Avail...

    2022

  20. [29]

    GREASE: Generate Factual and Counterfactual Explana- tions for GNN-based Recommendations,

    Z. Chen, F. Silvestri, J. Wang, Y . Zhang, Z. Huang, H. Ahn, and G. Tolomei, “GREASE: Generate Factual and Counterfactual Explana- tions for GNN-based Recommendations,” Aug. 2022, arXiv:2208.04222 [cs]. [Online]. Available: http://arxiv.org/abs/2208.04222

    2022 arXiv

  21. [30]

    Model-Agnostic Counterfactual Explanations of Recommendations,

    V . Kaffes, D. Sacharidis, and G. Giannopoulos, “Model-Agnostic Counterfactual Explanations of Recommendations,” in Proceedings of the 29th ACM Conference on User Modeling, Adaptation and Personalization. Utrecht Netherlands: ACM, Jun. 2021, pp. 280–285. [Online]. Available: h...

    2021

  22. [31]

    Shap-enhanced counterfactual explanations for recommendations,

    J. Zhong and E. Negre, “Shap-enhanced counterfactual explanations for recommendations,” in Proceedings of the 37th ACM/SIGAPP Symposium on Applied Computing , ser. SAC ’22. New York, NY , USA: Association for Computing Machinery, May 2022, pp. 1365–1372. [Online]. Available: h...

    2022

  23. [32]

    A survey on trustworthy recommender systems,

    Y . Ge, S. Liu, Z. Fu, J. Tan, Z. Li, S. Xu, Y . Li, Y . Xian, and Y . Zhang, “A survey on trustworthy recommender systems,” ACM Transactions on Recommender Systems, vol. 3, no. 2, pp. 1–68, 2024

    2024

  24. [33]

    Neural Collaborative Filtering,

    X. He, L. Liao, H. Zhang, L. Nie, X. Hu, and T.-S. Chua, “Neural Collaborative Filtering,” in Proceedings of the 26th International Conference on World Wide Web , ser. WWW ’17. Republic and Canton of Geneva, CHE: International World Wide Web Conferences Steering Committee, Apr...

    2017

  25. [34]

    Embarrassingly Shallow Autoencoders for Sparse Data,

    H. Steck, “Embarrassingly Shallow Autoencoders for Sparse Data,” in The World Wide Web Conference . San Francisco CA USA: ACM, May 2019, pp. 3251–3257. [Online]. Available: https://dl.acm.org/doi/ 10.1145/3308558.3313710

    2019

  26. [35]

    Deep Variational Autoencoder with Shallow Parallel Path for Top-N Recommendation (V ASP),

    V . Vanˇcura and P. Kord´ık, “Deep Variational Autoencoder with Shallow Parallel Path for Top-N Recommendation (V ASP),” in Artificial Neural Networks and Machine Learning – ICANN 2021 , I. Farka ˇs, P. Masulli, S. Otte, and S. Wermter, Eds. Cham: Springer International Publis...

    2021

  27. [36]

    Variational Autoencoders for Collaborative Filtering,

    D. Liang, R. G. Krishnan, M. D. Hoffman, and T. Jebara, “Variational Autoencoders for Collaborative Filtering,” in Proceedings of the 2018 World Wide Web Conference , ser. WWW ’18. Republic and Canton of Geneva, CHE: International World Wide Web Conferences Steering Committee,...

    2018

  28. [37]

    Strong mixed-integer programming formulations for trained neural networks,

    R. Anderson, J. Huchette, W. Ma, C. Tjandraatmadja, and J. P. Vielma, “Strong mixed-integer programming formulations for trained neural networks,” Mathematical Programming, vol. 183, no. 1, pp. 3–39, Sep

  29. [38]

    M. L. Bynum, G. A. Hackebeil, W. E. Hart, C. D. Laird, B. L. Nicholson, J. D. Siirola, J.-P. Watson, D. L. Woodruff, and others,Pyomo - Optimization Modeling in Python, 3rd Edition. Springer, 2021, vol. 67

    2021

  30. [39]

    SLIM: Sparse Linear Methods for Top-N Recommender Systems,

    X. Ning and G. Karypis, “SLIM: Sparse Linear Methods for Top-N Recommender Systems,” in 2011 IEEE 11th International Conference on Data Mining , Dec. 2011, pp. 497–506, iSSN: 2374-8486. [Online]. Available: https://ieeexplore.ieee.org/document/6137254

    2011

  31. [40]

    Enhancing V AEs for collaborative filtering: flexible priors & gating mechanisms,

    D. Kim and B. Suh, “Enhancing V AEs for collaborative filtering: flexible priors & gating mechanisms,” in Proceedings of the 13th ACM Conference on Recommender Systems, ser. RecSys ’19. New York, NY , USA: Association for Computing Machinery, Sep. 2019, pp. 403–407. [Online]. ...

    2019

  32. [41]

    Yelp Open Dataset

    Yelp, “Yelp Open Dataset.” [Online]. Available: https://business.yelp. com/data/resources/open-dataset/

  33. [42]

    Bridging language and items for retrieval and recommendation,

    Y . Hou, J. Li, Z. He, A. Yan, X. Chen, and J. McAuley, “Bridging language and items for retrieval and recommendation,” arXiv preprint arXiv:2403.03952, 2024

    2024 arXiv

  34. [43]

    The Netflix Prize,

    James Bennett and Stan Lanning, “The Netflix Prize,” Aug. 2007

    2007

  35. [44]

    Are We Explaining the Same Recommenders? Incorporating Recommender Performance for Evaluating Explainers,

    A. R. Mohammadi, A. Peintner, M. M ¨uller, and E. Zangerle, “Are We Explaining the Same Recommenders? Incorporating Recommender Performance for Evaluating Explainers,” inProceedings of the 18th ACM Conference on Recommender Systems, ser. RecSys ’24. New York, NY , USA: Associa...

    2024

  36. [45]

    Parallelizing the dual revised simplex method,

    Q. Huangfu and J. A. J. Hall, “Parallelizing the dual revised simplex method,” Mathematical Programming Computation , vol. 10, no. 1, pp. 119–142, Mar. 2018

    2018

  37. [46]

    Counterfactual Data-Augmented Sequential Recommendation,

    Z. Wang, J. Zhang, H. Xu, X. Chen, Y . Zhang, W. X. Zhao, and J.-R. Wen, “Counterfactual Data-Augmented Sequential Recommendation,” in Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval , ser. SIGIR ’21. New York, NY...

    2021

  38. [47]

    Counterfactual Collaborative Reasoning,

    J. Ji, Z. Li, S. Xu, M. Xiong, J. Tan, Y . Ge, H. Wang, and Y . Zhang, “Counterfactual Collaborative Reasoning,” in Proceedings of the Sixteenth ACM International Conference on Web Search and Data Mining, ser. WSDM ’23. New York, NY , USA: Association for Computing Machinery, ...

    2023

  39. [48]

    Biases in Recommendation System,

    S. Bhadani, “Biases in Recommendation System,” in Proceedings of the 15th ACM Conference on Recommender Systems , ser. RecSys ’21. New York, NY , USA: Association for Computing Machinery, Sep. 2021, pp. 855–859. [Online]. Available: https: //dl.acm.org/doi/10.1145/3460231.3473897

    2021

  40. [49]

    Enhancing Counterfactual Evaluation and Learning for Recommendation Systems,

    N. Felicioni, “Enhancing Counterfactual Evaluation and Learning for Recommendation Systems,” in Proceedings of the 16th ACM Conference on Recommender Systems , ser. RecSys ’22. New York, NY , USA: Association for Computing Machinery, Sep. 2022, pp. 739–741. [Online]. Available...

    2022

  41. [50]

    Counterfactual Explanation for Fairness in Recommendation,

    X. Wang, Q. Li, D. Yu, Q. Li, and G. Xu, “Counterfactual Explanation for Fairness in Recommendation,” ACM Trans. Inf. Syst. , vol. 42, no. 4, pp. 106:1–106:30, Mar. 2024. [Online]. Available: https://dl.acm.org/doi/10.1145/3643670

    2024 doi

  42. [2020]

    Available: https://doi.org/10.1007/s10107-020-01474-5

    [Online]. Available: https://doi.org/10.1007/s10107-020-01474-5

    doi

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.