Pith. sign in

REVIEW 2 major objections 6 minor 84 references

Synthetic tables that look statistically close still fail the SQL queries analysts actually run.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · grok-4.5

2026-07-11 22:58 UTC pith:O4CUHRPE

load-bearing objection Solid open benchmark that makes analytical SQL answers the evaluation object for synthetic tables; the five empirical patterns hold up under a large 49×11 campaign and a real stability check. the 2 major comments →

arxiv 2607.03926 v1 pith:O4CUHRPE submitted 2026-07-04 cs.DB cs.AI

TabQueryBench: A Query-Centric Benchmark for Synthetic Tabular Data

Jialin Zhang , Fenghao Dong , Yajie Zhou , Vyas Sekar , Shinan Liu This is my paper
classification cs.DB cs.AI
keywords synthetic tabular dataquery-centric fidelitySQL benchmarkstabular generative modelshigh-cardinality supporttail fidelityfidelity-cost tradeoff
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Synthetic tabular data is usually scored on how close its columns look to the real ones, or how well a machine-learning model trained on it performs. This paper argues that those scores miss the structure that matters for everyday analytics: the answers to the kinds of SQL questions people run on tables. The authors build TabQueryBench by distilling recurring analytical logic from public query collections into 44 reusable templates, then grounding those templates to each dataset so the same query families can be run fairly across many generators. On 49 datasets and 11 generators they show that even the strongest model reaches only about three-quarters of real-data query fidelity, with systematic collapse on high-cardinality discrete columns, local filtered slices versus global counterparts, and rare-event tails. The practical upshot is a clearer map of where current generators break and a cost-quality frontier in which a simple Bayesian network often wins for users who care about both answer quality and generation speed.

Core claim

Distance-based and ML-utility scores can make synthetic tables look faithful while the same tables still give wrong answers to analytical SQL. Across 49 datasets, RealTabFormer is the best query-centric model yet only reaches 0.75 ± 0.15 (real data = 1.00), and failures concentrate on high-cardinality discrete support, local conditional slices, and extreme tails.

What carries the argument

TabQueryBench: 44 reusable SQL-shaped query templates, taxonomized into five families (subgroup, conditional, tail/rarity, missingness, cardinality/range) from public analytical sources and grounded to each dataset by a policy-guided template-to-SQL pipeline that keeps queries schema-aware and comparable.

Load-bearing premise

The 44 templates drawn from the chosen public sources, once grounded by the fixed pipeline, form a fair and representative set of structural tests for the analytical uses the paper claims to cover.

What would settle it

If regenerated or independently authored query sets that still target the same five analytical families reverse the model ranking or erase the reported gaps on high-cardinality support, local slices, or extreme-tail recovery, the central claim that current generators systematically fail query-centric fidelity would not hold.

Watch this falsifier — get emailed when new claim-graph text bears on it.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 6 minor

Summary. TabQueryBench proposes a query-centric evaluation framework for synthetic tabular data: reusable SQL-shaped analytical queries act as structural assessors of fidelity rather than relying only on distance-based, privacy, or ML-utility metrics. From 12 public analytical-query sources the authors distill 44 templates in five families (subgroup, conditional, tail/rarity, missingness, cardinality/range), ground them to each of 49 datasets via a policy-guided template-to-SQL pipeline (deterministic profiling/binding/validation; LLM only in constrained realization), and evaluate 11 generative models. The main empirical claims are that (i) strong distance-based scores can coexist with substantially lower query-centric fidelity (best model RealTabFormer at 0.75±0.15 vs REAL=1.00), (ii) high-cardinality discrete support often collapses, (iii) local conditional slices are harder than global counterparts, (iv) tail fidelity degrades under stricter rarity thresholds, and (v) BayesNet offers the best fidelity–cost tradeoff on a common-9 runtime slice. A three-run ranking-stability study and open release of code, templates, and artifacts support the instrument.

Significance. The paper addresses a genuine and practically important gap: synthetic tabular data are often used for analytics, system testing, and data sharing, yet existing benchmarks rarely treat analytical query answers as first-class evaluation objects. The contribution is not a single theorem but a carefully constructed, extensible instrument with public provenance for templates, large multi-dataset/multi-model coverage, family-level diagnostics, a cost Pareto view, and an explicit stability audit of LLM-assisted grounding. If the reported patterns hold under the stated single-table mimicry scope—and the evidence is multi-faceted enough that they appear to—the work should influence both model selection practice and future generative-model design (e.g., high-cardinality support, rare-region, and local-slice objectives). Open code and artifacts further raise the work’s value as a community foundation.

major comments (2)
  1. The central numerical claims (e.g., RealTabFormer 0.75±0.15 query overall; local-slice drop of 0.11; ~40.7% rare-value recovery) depend on a precise definition of how a synthetic query answer is scored against the real answer. The main text and Appendix describe families, templates, and aggregate tables (Table 8, Figures 4–9) but do not give an explicit, auditable scoring map from (real result, synthetic result) to [0,1] per template type (counts, rates, rankings, support sets, range envelopes, missing rates). Please add a short formal definition (or algorithm box) covering result alignment, empty-support cases, and aggregation from template → family → overall, so that the headline numbers are independently checkable.
  2. Section 5.1 states that models were tuned within bounded search ranges, and free parameters include split ratio, synthetic row count, and family aggregation. For load-bearing ranking claims (RTF best; BayesNet best cost–fidelity), please report the final selected hyperparameters per model (or a compact appendix table) and state whether the overall query score is an unweighted mean of activated templates/families or a fixed weighted scheme. Without this, residual sensitivity of the reported orderings cannot be fully assessed even though the three-run stability study (Section 5.5) already bounds query-regeneration variance well.
minor comments (6)
  1. Figure 2 caption and surrounding text use abbreviated model names (T-DDPM, TPF, T-Syn) that are defined later in Table 2; define them at first use or move the abbreviation note earlier.
  2. Table 1 uses u for user-specified scale; a one-line legend note would help readers scanning the comparison table.
  3. In Finding 2 / Table 4, the prose sometimes cites slightly different distinct counts than the table (e.g., title 96,777 vs 96,779 elsewhere). Align the numbers for consistency.
  4. Section 3.2 scopes out long join chains and multi-table settings; a single sentence in the abstract or introduction stating the single-table focus would set expectations earlier for database readers.
  5. Appendix Table 9 heatmap uses TF for technical failure; ensure the main-text cost discussion (Figure 9, common-9) explicitly notes which models failed on which datasets so readers do not over-interpret missing cells.
  6. Minor typography: “ocurring” → “occurring” (Introduction); “pilcrow” artifacts in the author line of the source text should be cleaned in the camera-ready version.

Circularity Check

0 steps flagged

No significant circularity: empirical benchmark whose templates, models, and scores are externally sourced or held-out, not definitionally forced.

full rationale

TabQueryBench is a construction-and-measurement paper, not a first-principles derivation. Stage-1 templates are distilled from twelve named public sources (TPC-H/DS, ClickBench, H2O db-benchmark, RTABench, engine docs, SQL repos; Appendix A Table 7) and fixed before any model is run. Stage-2 grounding uses deterministic profiling/binding/validation; the LLM is confined to constrained realization and is re-run for stability (Section 5.5: Kendall W 0.927, Spearman 0.903). Generative models are third-party implementations trained only on the training split and scored against held-out real answers; REAL is the trivial self-comparison reference (score 1.00 by construction) used solely as an upper bound, not a fitted target. Distance-based metrics are classical (JSD/KS/TVD/Wasserstein) and reported separately from query scores. No equation equates a claimed prediction to a fitted input, no uniqueness theorem is imported from the authors, and no ansatz is smuggled via self-citation. The five reported patterns are multi-dataset, multi-model empirical outcomes under an explicit single-table mimicry scope. Author self-citations (network-trace papers) are peripheral and non-load-bearing. Residual design choices (template inventory, score aggregation) are ordinary for any new benchmark instrument and do not reduce the headline claims to their inputs by construction.

Axiom & Free-Parameter Ledger

5 free parameters · 4 axioms · 3 invented entities

The central empirical claims rest on design choices that define what counts as a fair query-centric assessor and on standard experimental knobs (splits, hyperparameter ranges, aggregation of family scores). No new physical entities are postulated; the invented objects are the benchmark artifacts themselves. Domain assumptions about what analytical structure matters are explicit design principles rather than hidden lemmas.

free parameters (5)
  • Train/eval split ratio and synthetic row count = 4:1; n_synth = n_eval
    4:1 real split; synthetic table sized to the evaluation split. Changes the absolute difficulty of rare-event and support recovery.
  • Per-model hyperparameter search ranges = model-specific ranges in §5.1
    Bounded grids for trees, epochs, diffusion steps, embedding sizes, etc., affect absolute fidelity and the reported Pareto frontier.
  • Tail rarity thresholds τ = 10%…0.1%
    Sweep from 10% to 0.1% defines the tail-degradation curves and the ~40.7% rare-value recovery headline.
  • Common-9 dataset slice for cost Pareto = C2,C7,C14,M4,M6,M8,N3,N6,N11
    Runtime–fidelity frontier is reported on a fixed nine-dataset subset; different slices could move which model is ‘best tradeoff’.
  • Query-score aggregation weights across families/templates = equal family aggregation as reported
    Overall query score averages family-level scores; reweighting tails vs missingness would change model orderings.
axioms (4)
  • domain assumption Answers to recurring analytical SQL patterns are valid primary structural assessors of synthetic tabular fidelity for sharing, testing, and analytics use cases.
    Stated as design Principle 1 and the core motivation in §§1–2; without it the benchmark measures a different quantity than claimed.
  • domain assumption The five families (subgroup, conditional, tail/rarity, missingness, cardinality/range) and 44 templates exhaust the transferable single-table analytical properties that are both recurring in public workloads and broadly groundable.
    Taxonomy construction criteria in §4.2; out-of-scope queries are listed but completeness is an assumption, not a proof.
  • ad hoc to paper Policy-constrained LLM template-to-SQL realization plus deterministic validation yields schema-aware queries that remain comparable across models and stable enough for ranking conclusions.
    Pipeline in §4.3 and stability study §5.5; three regenerations support ranking stability but do not prove invariance to other LLMs or policies.
  • domain assumption Standard distance metrics (JSD, KS, TVD, Wasserstein) and the authors’ query-answer similarity scores are meaningful higher-is-better / lower-is-better fidelity measures as normalized in the paper.
    Classical baseline appendix and evaluation sections; scoring details for every template family are partly deferred.
invented entities (3)
  • TabQueryBench template library (44 templates in five families) independent evidence
    purpose: Reusable analytical assessors that can be grounded across heterogeneous schemas.
    Constructed by the authors from 12 sources; independent evidence is the public provenance of sources and open release, not an external physical measurement.
  • Policy-guided template-to-SQL grounding pipeline independent evidence
    purpose: Make queries schema-aware while preserving cross-model comparability.
    Core methodological invention of the paper; falsifiable via the released code and stability experiments.
  • Query-centric fidelity scores (overall and per-family) no independent evidence
    purpose: Scalar and diagnostic measures of whether synthetic tables preserve analytical query answers.
    Defined by the benchmark; external validity depends on the domain axioms above.

pith-pipeline@v1.1.0-grok45 · 35586 in / 3501 out tokens · 38424 ms · 2026-07-11T22:58:16.353808+00:00 · methodology

0 comments
read the original abstract

Synthetic tabular data support use cases like data sharing, model development under access restrictions, and rapid prototyping of analytical workflows. Modern generative models are evaluated by their statistical similarity, correlation structure, privacy, and downstream machine-learning utility. However, such evaluations leave a gap: they rarely test the structure that matters for analytical queries. We present TabQueryBench, a query-centric benchmark that uses SQL-shaped analytical queries as structural assessors for synthetic data fidelity. It provides an extensible foundation for query-centric synthetic-data evaluation. From 12 public sources of analytical queries, TabQueryBench taxonomizes recurring cross-domain logic into 44 reusable query templates and grounds them to each dataset via a policy-guided template-to-SQL pipeline. This makes queries schema-aware while preserving comparability across generative models. Across 49 datasets and 11 generative models, it activates 10-12 templates per dataset, producing more than 100 executable SQL queries per dataset. Our systematic experiments show five main patterns. First, current tabular generative models can have good distance-based fidelity, but they still fall short on query-centric fidelity: RealTabFormer achieves the highest query-centric fidelity, but it only reaches 0.75 +/- 0.15 (REAL data score is 1.00). Second, tabular generative models struggle with very high-cardinality discrete support. Third, SOTA generative models preserve good global conditional query-centric fidelity, but fail more on local queries. Fourth, tail fidelity deteriorates as queries move toward the extreme tail; even the best model recovers only about 40.7% of real rare values. Finally, there is a fidelity-cost tradeoff in tabular generation: BayesNet offers the strongest tradeoff, with slightly lower query-centric fidelity but much lower generation cost.

Figures

Figures reproduced from arXiv: 2607.03926 by Fenghao Dong, Jialin Zhang, Shinan Liu, Vyas Sekar, Yajie Zhou.

Figure 1
Figure 1. Figure 1: Overview of TabQueryBench, including the benchmark design and representative evaluation results. preserves schema-level realism while enabling comparable evalu￾ation across datasets, domains, and generative models ( [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Comparison between query-centric fidelity vs. dis [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: The Query Template Taxonomy. a privacy budget; conditional generation under task-specific con￾straints, where query-centric fidelity is not the primary target; or text-to-table generation from natural-language prompts. The gen￾erative models included in the current roster all fall within this mimicry-style setting. 3.3 Dataset Suite TabQueryBench curates 49 datasets organized by feature regime: 19 categori… view at source ↗
Figure 4
Figure 4. Figure 4: Distance-based and query overall scores do not [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗
Figure 7
Figure 7. Figure 7: Tail overall, coverage, and size under progressively [PITH_FULL_IMAGE:figures/full_fig_p010_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Tail degradation across individual generative mod [PITH_FULL_IMAGE:figures/full_fig_p011_8.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

84 extracted references · 9 canonical work pages

  1. [1]

    Altinity. [n.d.]. ClickHouse TPC-DS Benchmark. https://github.com/Altinity/tpc- ds. https://github.com/Altinity/tpc-ds

  2. [2]

    [n.d.].All Druid SQL Functions

    Apache Druid. [n.d.].All Druid SQL Functions. https://druid.apache.org/docs/ latest/querying/sql-functions/

  3. [3]

    Apellániz, Ana Jiménez, Borja Arroyo Galende, Juan Parras, and Santiago Zazo

    Patricia A. Apellániz, Ana Jiménez, Borja Arroyo Galende, Juan Parras, and Santiago Zazo. 2024. Synthetic Tabular Data Validation: A Divergence-Based Ap- proach.IEEE Access12 (2024), 103895–103907. https://doi.org/10.1109/ACCESS. 2024.3434582

    doi:10.1109/access 2024

  4. [4]

    arka420. [n.d.]. Insurance-Cost-Project-Using-SQL. https://github.com/arka420/ Insurance-Cost-Project-Using-SQL. https://github.com/arka420/Insurance- Cost-Project-Using-SQL

  5. [5]

    Bates, Iva Špakulová, Iain Dove, and Andrew Mealor

    Andrew G. Bates, Iva Špakulová, Iain Dove, and Andrew Mealor. 2019.ONS Methodology Working Paper Series Number 16: Synthetic Data Pilot. Technical Report. Office for National Statistics. https://www.ons.gov.uk/methodology/ methodologicalpublications/generalmethodology/onsworkingpaperseries/ onsmethodologyworkingpaperseriesnumber16syntheticdatapilot Publis...

    2019

  6. [6]

    Mantovani, Jan N

    Bernd Bischl, Giuseppe Casalicchio, Matthias Feurer, Pieter Gijsbers, Frank Hut- ter, Michel Lang, Rafael G. Mantovani, Jan N. van Rijn, and Joaquin Vanschoren

  7. [7]

    InProceedings of the Neural Information Processing Systems Track on Datasets and Benchmarks, Vol

    OpenML Benchmarking Suites. InProceedings of the Neural Information Processing Systems Track on Datasets and Benchmarks, Vol. 1

  8. [8]

    Hasan Alp Caferoğlu, Mehmet Serhat Çelik, and Özgür Ulusoy. 2025. SING-SQL: A Synthetic Data Generation Framework for In-Domain Text-to-SQL Translation. https://doi.org/10.48550/arXiv.2509.25672 arXiv:2509.25672 [cs.AI]

    doi:10.48550/arxiv.2509.25672 2025

  9. [9]

    Andrew Chu, Xi Jiang, Shinan Liu, Arjun Bhagoji, Francesco Bronzino, Paul Schmitt, and Nick Feamster. 2024. Feasibility of state space models for network traffic generation. InProceedings of the 2024 SIGCOMM Workshop on Networks for AI Computing. 9–17

    2024

  10. [10]

    Andrew Chu, Xi Jiang, Shinan Liu, Arjun Bhagoji, Francesco Bronzino, Paul Schmitt, and Nick Feamster. 2026. Netssm: Multi-flow and state-aware network trace generation using state-space models.Proceedings of the ACM on Networking 4, CoNEXT1 (2026), 1–24

    2026

  11. [11]

    [n.d.].Aggregate Functions

    ClickHouse. [n.d.].Aggregate Functions. https://clickhouse.com/docs/sql- reference/aggregate-functions/reference

  12. [12]

    ClickHouse. 2022. ClickBench: A Benchmark for Analytical Databases. https://github.com/ClickHouse/ClickBench. https://github.com/ClickHouse/ ClickBench GitHub benchmark

    2022

  13. [13]

    CWI Database Architectures Group. [n.d.]. Reproducing TPC-DS Qualification Results. https://github.com/cwida/tpcds-result-reproduction. https://github. com/cwida/tpcds-result-reproduction

  14. [14]

    DataCebo. 2026. Welcome to SDGym! https://docs.sdv.dev/sdgym Official documentation

    2026

  15. [15]

    Narasayya

    Bailu Ding, Surajit Chaudhuri, Johannes Gehrke, and Vivek R. Narasayya. 2021. DSB: A Decision Support Benchmark for Workload-Driven and Traditional Database Systems.Proceedings of the VLDB Endowment14, 13 (2021), 3376–3388. https://doi.org/10.14778/3484224.3484234

    doi:10.14778/3484224.3484234 2021

  16. [16]

    Junhui Ding, Xinchen Zhang, Xiaohui Xie, and Shinan Liu. 2026. TraceCodec: A Compiler-Backed Neural Codec for Stateful Multi-Flow Network Traffic Traces. arXiv preprint arXiv:2605.29941(2026)

    Pith/arXiv arXiv 2026

  17. [17]

    Yuntao Du and Ninghui Li. 2025. Systematic Assessment of Tabular Data Syn- thesis. InProceedings of the 2025 ACM SIGSAC Conference on Computer and Communications Security. 2414–2428. https://doi.org/10.1145/3719027.3765067

    doi:10.1145/3719027.3765067 2025

  18. [18]

    Khaled El Emam, Lucy Mosquera, Xi Fang, and Alaa El-Hussuna. 2022. Utility Metrics for Evaluating Synthetic Health Data Generation Methods: Validation Study.JMIR Medical Informatics10, 4 (2022), e35734. https://doi.org/10.2196/ 35734

    2022

  19. [19]

    Nick Erickson, Lennart Purucker, Andrej Tschalzev, David Holzmüller, Pra- teek Mutalik Desai, David Salinas, and Frank Hutter. 2026. TabArena: A Living Benchmark for Machine Learning on Tabular Data. InThe Thirty-ninth Annual Conference on Neural Information Processing Systems Datasets and Benchmarks Track. https://openreview.net/forum?id=jZqCqpCLdU

    2026

  20. [20]

    [n.d.].GoogleSQL Approximate Aggregate Func- tions

    Google Cloud. [n.d.].GoogleSQL Approximate Aggregate Func- tions. https://cloud.google.com/bigquery/docs/reference/standard- sql/approximate_aggregate_functions

  21. [21]

    Léo Grinsztajn, Edouard Oyallon, and Gaël Varoquaux. 2022. Why Do Tree-Based Models Still Outperform Deep Learning on Typical Tabular Data?. InAdvances in Neural Information Processing Systems, Vol. 35. 507–520

    2022

  22. [22]

    Ragini Gupta, Shinan Liu, Ruixiao Zhang, Xinyue Hu, Xiaoyang Wang, Hadjer Benkraouda, Pranav Kommaraju, Phuong Cao, Nick Feamster, and Klara Nahrst- edt. 2025. Generative active adaptation for drifting and imbalanced network intrusion detection.arXiv preprint arXiv:2503.03022(2025)

    Pith/arXiv arXiv 2025

  23. [23]

    Andrés Guzmán-Cordero, Floor Eijkelboom, and Jan-Willem van de Meent. 2025. Exponential Family Variational Flow Matching for Tabular Data Generation. In Forty-second International Conference on Machine Learning. https://openreview. net/forum?id=kjtvCSkSsy

    2025

  24. [24]

    H2O.ai. 2020. Database-like Operations Benchmark. https://h2oai.github.io/db- benchmark/. https://h2oai.github.io/db-benchmark/ Benchmark report and reproducibility repository

    2020

  25. [25]

    Lasse Hansen, Nabeel Seedat, Mihaela van der Schaar, and Andrija Petrovic

  26. [26]

    InAdvances in Neural Information Pro- cessing Systems, Vol

    Reimagining Synthetic Tabular Data Generation through Data-Centric AI: A Comprehensive Benchmark. InAdvances in Neural Information Pro- cessing Systems, Vol. 36. https://papers.nips.cc/paper_files/paper/2023/hash/ 6aa9a05b929fb08ff46a58cab6cf860d-Abstract-Datasets_and_Benchmarks.html Datasets and Benchmarks Track

    2023

  27. [27]

    Mikel Hernandez, Gorka Epelde, Ane Alberdi, Rodrigo Cilla, and Debbie Rankin

  28. [28]

    https://doi.org/10.1055/s-0042-1760247

    Synthetic Tabular Data Evaluation in the Health Domain Covering Resem- blance, Utility, and Privacy Dimensions.Methods of Information in Medicine62, S 01 (2023), e19–e38. https://doi.org/10.1055/s-0042-1760247

    doi:10.1055/s-0042-1760247 2023

  29. [29]

    Julian Höllig and Michaela Geierhos. 2025. Utility Meets Privacy: A Critical Evaluation of Tabular Data Synthesizers.IEEE Access13 (2025), 44497–44509. https://doi.org/10.1109/ACCESS.2025.3549680

    work page doi:10.1109/access.2025.3549680 2025

  30. [30]

    Valter Hudovernik, Martin Jurković, and Erik Štrumbelj. 2024. Benchmarking the Fidelity and Utility of Synthetic Relational Data.arXiv preprint arXiv:2410.03411 (2024). https://doi.org/10.48550/arXiv.2410.03411

    Pith/arXiv arXiv doi:10.48550/arxiv.2410.03411 2024

  31. [31]

    Hugging Face. [n.d.]. Hugging Face Datasets. https://huggingface.co/datasets. https://huggingface.co/datasets Accessed: 2026-07-01

    2026

  32. [32]

    Tobias Hyrup, Anton Danholt Lautrup, Arthur Zimek, and Peter Schneider-Kamp

  33. [33]

    https://doi.org/10.1016/j.mlwa.2024.100608

    Sharing is CAIRing: Characterizing Principles and Assessing Properties of Universal Privacy Evaluation for Synthetic Tabular Data.Machine Learning with Applications18 (2024), 100608. https://doi.org/10.1016/j.mlwa.2024.100608

    work page doi:10.1016/j.mlwa.2024.100608 2024

  34. [34]

    Xi Jiang, Shinan Liu, Aaron Gember-Jacobson, Arjun Nitin Bhagoji, Paul Schmitt, Francesco Bronzino, and Nick Feamster. 2024. Netdiffusion: Network data aug- mentation through protocol-constrained traffic generation.Proceedings of the ACM on Measurement and Analysis of Computing Systems8, 1 (2024), 1–32

    2024

  35. [35]

    Xi Jiang, Shinan Liu, Aaron Gember-Jacobson, Paul Schmitt, Francesco Bronzino, and Nick Feamster. 2023. Generative, high-fidelity network traces. InProceedings of the 22nd ACM Workshop on Hot Topics in Networks. 131–138

    2023

  36. [36]

    Xiangjian Jiang, Nikola Simidjievski, and Mateja Jamnik. 2026. TabStruct: Measur- ing Structural Fidelity of Tabular Data. InThe Fourteenth International Conference on Learning Representations. https://openreview.net/forum?id=XOPH34Extq Oral presentation

    2026

  37. [37]

    Alexia Jolicoeur-Martineau, Kilian Fatras, and Tal Kachman. 2024. Generating and Imputing Tabular Data via Diffusion and Flow-based Gradient-Boosted Trees. InProceedings of The 27th International Conference on Artificial Intelligence and Statistics (Proceedings of Machine Learning Research), Sanjoy Dasgupta, Stephan Mandt, and Yingzhen Li (Eds.), Vol. 238...

    2024

  38. [38]

    Bayrem Kaabachi, Jérémie Despraz, Thierry Meurers, Karen Otte, Mehmed Halilovic, Bogdan Kulynych, Fabian Prasser, and Jean Louis Raisaro. 2025. A Scoping Review of Privacy and Utility Metrics in Medical Synthetic Data.npj Digital Medicine8, 1 (2025), 60. https://doi.org/10.1038/s41746-024-01359-3

    work page doi:10.1038/s41746-024-01359-3 2025

  39. [39]

    Kaggle. [n.d.]. Kaggle. https://www.kaggle.com. https://www.kaggle.com Accessed: 2026-07-01

    2026

  40. [40]

    Shivasankari Kannan, Yeounoh Chung, Amita Gondi, Tristan Swadell, and Fatma Ozcan. 2025. High-Fidelity And Complex Test Data Generation For Google SQL Code Generation Services. https://doi.org/10.48550/arXiv.2504.17203 arXiv:2504.17203 [cs.DB]

    work page doi:10.48550/arxiv.2504.17203 2025

  41. [41]

    Markelle Kelly, Rachel Longjohn, and Kolby Nottingham. [n.d.]. The UCI Machine Learning Repository. https://archive.ics.uci.edu. https://archive.ics.uci.edu Accessed: 2026-07-01

    2026

  42. [42]

    Akim Kotelnikov, Dmitry Baranchuk, Ivan Rubachev, and Artem Babenko. 2023. TabDDPM: Modelling Tabular Data with Diffusion Models. InProceedings of the 40th International Conference on Machine Learning (Proceedings of Machine Learning Research), Vol. 202. 17564–17579. https://proceedings.mlr.press/v202/ kotelnikov23a.html

    2023

  43. [43]

    Anton D Lautrup, Tobias Hyrup, Arthur Zimek, and Peter Schneider-Kamp

  44. [44]

    https://doi.org/10.1007/s10618-024-01081-4

    Syntheval: a framework for detailed utility and privacy evaluation of tabular synthetic data.Data Mining and Knowledge Discovery39, 1 (2025), 6. https://doi.org/10.1007/s10618-024-01081-4

    doi:10.1007/s10618-024-01081-4 2025

  45. [45]

    Junwei Ma, Apoorv Dankar, George Stein, Guangwei Yu, and Anthony Caterini

  46. [46]

    InNeurIPS 2023 Second Table Representation Learning Workshop

    TabPFGen – Tabular Data Generation with TabPFN. InNeurIPS 2023 Second Table Representation Learning Workshop. https://openreview.net/forum?id= 4MkkNsAEmO

    2023

  47. [47]

    Kirkpatrick, D

    Mark Mazumder, Colby Banbury, Xiaozhe Yao, Bojan Karlaš, William Gaviria Ro- jas, Sudnya Diamos, Greg Diamos, Lynn He, Alicia Parrish, Hannah Rose Kirk, Jessica Quaye, Charvi Rastogi, Douwe Kiela, David Jurado, David Kanter, Rafael Mosquera, Will Cukierski, Juan Ciro, Lora Aroyo, Bilge Acun, Lingjiao Chen, Mehul Raje, Max Bartolo, Evan Sabri Eyuboglu, Ami...

    2023

  48. [48]

    Raghunath Othayoth Nambiar and Meikel Poess. 2006. The Making of TPC-DS. InProceedings of the 32nd International Conference on Very Large Data Bases. VLDB Endowment, 1049–1058

    2006

  49. [49]

    Noshin Nawar Nawar. [n.d.]. Exploratory Analysis of Car Evaluation Dataset with SQL. https://github.com/nehanawar025/Exploratory-Analysis-of-Car- Evaluation-Dataset-with-SQL. https://github.com/nehanawar025/Exploratory- Analysis-of-Car-Evaluation-Dataset-with-SQL

  50. [50]

    OpenML. [n.d.]. OpenML. https://www.openml.org. https://www.openml.org Accessed: 2026-07-01

    2026

  51. [51]

    Neha Patki, Roy Wedge, and Kalyan Veeramachaneni. 2016. The Synthetic Data Vault. In2016 IEEE International Conference on Data Science and Advanced Analytics. 399–410. https://doi.org/10.1109/DSAA.2016.49

    doi:10.1109/dsaa.2016.49 2016

  52. [52]

    Personal Data Protection Commission Singapore. 2024. Proposed Guide on Synthetic Data Generation. Guidance document. https: //www.pdpc.gov.sg/organisations/resources/guidance-by-topic/proposed- guide-on-synthetic-data-generation Published 24 September 2024

    2024

  53. [53]

    Haoyue Ping, Julia Stoyanovich, and Bill Howe. 2017. DataSynthesizer: Privacy- Preserving Synthetic Datasets. InProceedings of the 29th International Conference on Scientific and Statistical Database Management (SSDBM ’17). 42:1–42:5. https: //doi.org/10.1145/3085504.3091117

    work page doi:10.1145/3085504.3091117 2017

  54. [54]

    Meikel Poess and Chris Floyd. 2000. New TPC Benchmarks for Decision Support and Web Commerce.ACM SIGMOD Record29, 4 (2000), 64–71. https://doi.org/ 10.1145/369275.369291

    work page doi:10.1145/369275.369291 2000

  55. [55]

    Meikel Poess, Raghunath Othayoth Nambiar, and David Walrath. 2007. Why You Should Run TPC-DS: A Workload Analysis. InProceedings of the 33rd International Conference on Very Large Data Bases. 1138–1149. https://www.vldb.org/conf/ 2007/papers/industrial/p1138-poess.pdf

    2007

  56. [56]

    Meikel Poess, Bryan Smith, Lubor Kollár, and Per-Åke Larson. 2002. TPC-DS, Taking Decision Support Benchmarking to the Next Level. InProceedings of the 2002 ACM SIGMOD International Conference on Management of Data. 582–587. https://doi.org/10.1145/564691.564759

    work page doi:10.1145/564691.564759 2002

  57. [57]

    Zhaozhi Qian, Rob Davis, and Mihaela van der Schaar. 2023. Syn- thcity: A Benchmark Framework for Diverse Use Cases of Tabular Synthetic Data. InAdvances in Neural Information Processing Sys- tems, Vol. 36. https://papers.nips.cc/paper_files/paper/2023/hash/ 09723c9f291f6056fd1885081859c186-Abstract-Datasets_and_Benchmarks.html Datasets and Benchmarks Track

    2023

  58. [58]

    Gabriele Santangelo, Giovanna Nicora, Riccardo Bellazzi, and Arianna Dagliati

  59. [59]

    https://doi.org/10.1186/s12911-024-02731-9

    How Good Is Your Synthetic Data? SynthRO, a Dashboard to Evaluate and Benchmark Synthetic Tabular Data.BMC Medical Informatics and Decision Making25, 1 (2025), 89. https://doi.org/10.1186/s12911-024-02731-9

    doi:10.1186/s12911-024-02731-9 2025

  60. [60]

    Aniruddha Sen, Christine Task, Dhruv Kapur, Gary Howarth, and Karan Bhagat

  61. [61]

    Advances in Neural Information Processing Systems36 (2023), 51409–51420

    Diverse Community Data for Benchmarking Data Privacy Algorithms. Advances in Neural Information Processing Systems36 (2023), 51409–51420

    2023

  62. [62]

    Shagufta-DataAnalyst. [n.d.]. insurance-sql-analysis. https://github.com/ Shagufta-DataAnalyst/insurance-sql-analysis. https://github.com/Shagufta- DataAnalyst/insurance-sql-analysis

  63. [63]

    Juntong Shi, Minkai Xu, Harper Hua, Hengrui Zhang, Stefano Ermon, and Jure Leskovec. 2025. TabDiff: a Mixed-type Diffusion Model for Tabular Data Gen- eration. InThe Thirteenth International Conference on Learning Representations. https://openreview.net/forum?id=swvURjrt8z

    2025

  64. [64]

    Andrey Sidorenko, Michael Platzer, Mario Scriminaci, and Paul Tiwald. 2025. Benchmarking Synthetic Tabular Data: A Multi-Dimensional Evaluation Frame- work.arXiv preprint arXiv:2504.01908(2025). https://doi.org/10.48550/arXiv. 2504.01908

    Pith/arXiv arXiv doi:10.48550/arxiv 2025

  65. [65]

    Raab, Beata Nowok, Chris Dibben, and Aleksandra Slavković

    Joshua Snoke, Gillian M. Raab, Beata Nowok, Chris Dibben, and Aleksandra Slavković. 2018. General and Specific Utility Measures for Synthetic Data.Journal of the Royal Statistical Society Series A: Statistics in Society181, 3 (2018), 663–688. https://doi.org/10.1111/rssa.12358

    work page doi:10.1111/rssa.12358 2018

  66. [66]

    Snowflake Inc. 2026. PERCENTILE_CONT — Snowflake Documentation. https: //docs.snowflake.com/en/sql-reference/functions/percentile_cont. https://docs. snowflake.com/en/sql-reference/functions/percentile_cont

    2026

  67. [67]

    Snowflake Inc. 2026. WIDTH_BUCKET — Snowflake Documentation. https: //docs.snowflake.com/en/sql-reference/functions/width_bucket. https://docs. snowflake.com/en/sql-reference/functions/width_bucket

    2026

  68. [68]

    Solatorio and Olivier Dupriez

    Aivin V. Solatorio and Olivier Dupriez. 2023. REaLTabFormer: Generating Realistic Relational and Tabular Data using Transformers.arXiv preprint arXiv:2302.02041(2023). https://doi.org/10.48550/arXiv.2302.02041

    Pith/arXiv arXiv doi:10.48550/arxiv.2302.02041 2023

  69. [69]

    Yuchao Tao, Ryan McKenna, Michael Hay, Ashwin Machanavajjhala, and Gerome Miklau. 2021. Benchmarking Differentially Private Synthetic Data Generation Algorithms.arXiv preprint arXiv:2112.09238(2021). https://doi.org/10.48550/ arXiv.2112.09238

    Pith/arXiv arXiv 2021

  70. [70]

    Christine Task, Karan Bhagat, and Gary Howarth. 2023. SDNist v2: Deidentified Data Report Tool. https://doi.org/10.18434/MDS2-2943

    doi:10.18434/mds2-2943 2023

  71. [71]

    Timescale. 2025. RTABench: A Benchmark for Real-Time Analytics Applications. https://github.com/timescale/rtabench. https://github.com/timescale/rtabench GitHub benchmark

    2025

  72. [72]

    Van Tran, Shinan Liu, Tian Li, and Nick Feamster. 2025. Quantifying the Pri- vacy Implications of High-Fidelity Synthetic Network Traffic.arXiv preprint arXiv:2511.20497(2025)

    arXiv 2025

  73. [73]

    Transaction Processing Performance Council. [n.d.]. TPC Benchmark DS (TPC- DS) Standard Specification. https://www.tpc.org/tpcds/. https://www.tpc.org/ tpcds/

  74. [74]

    Transaction Processing Performance Council. [n.d.]. TPC Benchmark H (TPC-H) Standard Specification. https://www.tpc.org/tpch/. https://www.tpc.org/tpch/

  75. [75]

    Trino Software Foundation. 2026. Aggregate Functions — Trino Documentation. https://trino.io/docs/current/functions/aggregate.html. https://trino.io/docs/ current/functions/aggregate.html

    2026

  76. [76]

    Boris van Breugel, Zhaozhi Qian, and Mihaela van der Schaar. 2023. Synthetic Data, Real Errors: How (Not) to Publish and Use Synthetic Data. InProceedings of the 40th International Conference on Machine Learning (Proceedings of Machine Learning Research), Vol. 202. 34793–34808. https://proceedings.mlr.press/v202/ van-breugel23a.html

    2023

  77. [77]

    Jason Walonoski, Mark Kramer, Joseph Nichols, Andre Quina, Chris Moesel, Dylan Hall, Carlton Duffett, Kudakwashe Dube, Thomas Gallagher, and Scott McLachlan. 2018. Synthea: An Approach, Method, and Software Mechanism for Generating Synthetic Patients and the Synthetic Electronic Health Care Record. Journal of the American Medical Informatics Association25...

    doi:10.1093/jamia/ocx079 2018

  78. [78]

    Xinyue Wang, Hafiz Asif, and Jaideep Vaidya. 2023. Preserving Missing Data Distribution in Synthetic Data. InProceedings of the ACM Web Conference 2023. 2110–2121. https://doi.org/10.1145/3543507.3583297

    work page doi:10.1145/3543507.3583297 2023

  79. [79]

    Watson, Kristin Blesch, Jan Kapar, and Marvin N

    David S. Watson, Kristin Blesch, Jan Kapar, and Marvin N. Wright. 2023. Ad- versarial Random Forests for Density Estimation and Generative Modeling. InProceedings of the 26th International Conference on Artificial Intelligence and Statistics (Proceedings of Machine Learning Research), Vol. 206. 5357–5375. https://proceedings.mlr.press/v206/watson23a.html

    2023

  80. [80]

    Lei Xu, Maria Skoularidou, Alfredo Cuesta-Infante, and Kalyan Veeramachaneni

Showing first 80 references.