pith. sign in

arxiv: 2605.23887 · v1 · pith:VKY772L6new · submitted 2026-05-22 · 💻 cs.DB · cs.AI· cs.CR· cs.LG· cs.MA

CHRONOS: Temporally-Aware Multi-Agent Coordination for Evolving Data Marketplaces

Joydeep Chandra This is my paper

Pith reviewed 2026-05-25 02:12 UTC · model grok-4.3

classification 💻 cs.DB cs.AIcs.CRcs.LGcs.MA
keywords data marketplacestemporal knowledge graphsdifferential privacyShapley valuationmulti-agent coordinationneural ODEchangepoint detectionhybrid indexes
0
0 comments X

The pith

CHRONOS is a three-layer architecture that jointly addresses stale hybrid indexes, misattributed Shapley pricing after shifts, and over-consumption of a shared differential-privacy budget in temporal knowledge-graph data marketplaces.

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

The paper establishes that static marketplace designs fail in three coupled ways as data evolves: shortcut edges become stale and lower recall, stationary pricing misattributes value after distribution changes, and uncoordinated agents exhaust a shared privacy budget. CHRONOS supplies a unified treatment through explicit public-private separation. Layer one uses neural-ODE temporal decay on shortcut edges to bound per-query recall loss. Layer two conditions Shapley valuations on detected changepoints to control error under noise. Layer three applies EXP3-IX for sublinear regret while releasing a privatized affinity matrix each epoch under the Gaussian mechanism, with all retrieval and ranking treated as post-processing. The reported operating point reaches 0.937 recall at ten, 2.74 queries per second, 161 ms latency, and total epsilon of 4.25 at delta 10 to the power of negative 6. A sympathetic reader cares because the design shows how public index routing and low-sensitivity scheduling can still deliver utility even when privatized valuations remain noise-dominated.

Core claim

CHRONOS applies neural-ODE temporal decay to shortcut edges, providing a per-query expected recall-loss bound of O(Pq λ δt) together with a monotone-envelope guarantee that reduces bound looseness to 1.8 to 3.2 times observed loss; conditions Shapley valuation on detected changepoints and supplies finite-sample error guarantees under noise; and employs EXP3-IX to achieve O(sqrt(T log T)) regret while enforcing ε and δ differential privacy via moments accounting, releasing a privatized affinity matrix per epoch with the Gaussian mechanism so that all retrieval and ranking incur no extra privacy cost. Across four benchmarks the system records 0.937 recall at ten, 2.74 queries per second, 161ms

What carries the argument

Three-layer architecture with explicit public-private separation in which layer one performs neural-ODE temporal decay on shortcut edges, layer two performs changepoint-conditioned Shapley valuation, and layer three performs EXP3-IX coordination with Gaussian-mechanism release of the affinity matrix.

Load-bearing premise

The monotone-envelope guarantee that reduces bound looseness to 1.8 to 3.2 times observed loss together with the finite-sample error guarantees for changepoint-conditioned Shapley valuation under noise.

What would settle it

An experiment in which, after a distribution shift, either the observed recall loss exceeds 3.2 times the per-query bound or the error in the released Shapley valuations exceeds the finite-sample guarantee provided by the changepoint conditioning.

Figures

Figures reproduced from arXiv: 2605.23887 by Joydeep Chandra.

Figure 1
Figure 1. Figure 1: Public/private data flow under the trusted-curator [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Chronos architecture. 4.2 Layer 2: Event-Conditioned MPV Real-world events alter dataset marginal values in ways that static Shapley misses [28]. EC-MPV addresses this by conditioning on detected distributional changepoints. Definition 4 (Temporal KG Affinity). The temporal affinity affKG(𝑢, 𝑣, 𝑡) = decay(𝑡 − 𝑡𝑒 ) · affstatic (𝑢, 𝑣) where 𝑡𝑒 is the creation timestamp and affstatic (𝑢, 𝑣) ∈ [0, 1] is Louvai… view at source ↗
Figure 3
Figure 3. Figure 3: Recall@10 degradation on MIMIC-IV (𝜆 ≈ 12 changes/day/shortcut) over 90 simulated days without re￾indexing [PITH_FULL_IMAGE:figures/full_fig_p011_3.png] view at source ↗
read the original abstract

Temporal knowledge-graph data marketplaces face three coupled failures in static designs: stale hybrid index shortcuts reduce recall as edges evolve, stationary Shapley pricing misattributes value after distribution shifts, and uncoordinated agents over-consume a shared differential-privacy budget. We present CHRONOS, a three-layer architecture providing a unified treatment of these challenges with explicit public and private separation. Layer one applies neural-ODE temporal decay to shortcut edges, providing a per-query expected recall-loss bound of Big-O of Pq lambda delta t, with a monotone-envelope guarantee reducing bound looseness to 1.8 to 3.2 times observed loss. Layer two conditions Shapley valuation on detected changepoints and provides finite-sample error guarantees under noise. Layer three uses EXP3-IX to achieve Big-O of the square root of T log T regret while enforcing epsilon and delta differential privacy via moments accounting. CHRONOS releases a privatized affinity matrix per epoch using the Gaussian mechanism; all retrieval and ranking are post-processing, incurring no extra privacy cost. We provide multi-epoch settlement, scalability analysis for 500 sellers, and comparisons against accelerated baselines. Across four benchmarks, CHRONOS shows 0.937 recall at ten, 2.74 queries per second, 161 ms latency, and total epsilon of 4.25 at delta of 10 to the power of negative 6 under zCDP composition. These results indicate a competitive operating point. A limitation is that at this privacy level, released valuations remain noise-dominated; utility derives primarily from public index routing and adaptive scheduling driven by low-sensitivity statistics.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. CHRONOS presents a three-layer architecture for temporally-aware multi-agent coordination in evolving data marketplaces. Layer 1 uses neural-ODE temporal decay on shortcut edges to bound per-query expected recall loss by O(Pq λ Δt) with a monotone-envelope guarantee claimed to reduce looseness to 1.8–3.2× observed loss. Layer 2 conditions Shapley valuation on detected changepoints and supplies finite-sample error guarantees under noise. Layer 3 applies EXP3-IX for O(√(T log T)) regret while enforcing (ε,δ)-DP via moments accounting and the Gaussian mechanism on the affinity matrix (all retrieval/ranking is post-processing). The paper reports multi-epoch settlement, scalability to 500 sellers, and empirical results across four benchmarks: 0.937 recall@10, 2.74 qps, 161 ms latency, and total ε=4.25 at δ=10^{-6} under zCDP composition, while noting that valuations remain noise-dominated at this privacy level.

Significance. If the stated recall-loss bound and finite-sample Shapley guarantees survive rigorous derivation and noise propagation, the unified treatment of temporal index staleness, post-shift value attribution, and coordinated privacy-budget consumption would be a meaningful contribution to the design of dynamic data marketplaces. The reported operating point is competitive with accelerated baselines and the explicit public/private separation is a useful architectural choice; however, the absence of the supporting derivations for the two load-bearing theoretical claims limits the strength of the significance assessment.

major comments (2)
  1. [Abstract / Layer one] Abstract / Layer one: the monotone-envelope guarantee that reduces the O(Pq λ Δt) recall-loss bound looseness to 1.8–3.2× observed loss is asserted without derivation, statement of the monotonicity conditions, or proof that the envelope survives the distribution shifts the system is designed to handle; this is load-bearing for justifying the hybrid-index shortcut.
  2. [Abstract / Layer two] Abstract / Layer two: the finite-sample error guarantees for changepoint-conditioned Shapley valuation are stated to hold even after Gaussian noise is added to the affinity matrix, yet no explicit propagation of that noise through the estimator is supplied; without it the guarantee does not automatically survive the privacy mechanism.
minor comments (2)
  1. [Abstract] The paper already notes that released valuations remain noise-dominated; a short additional sentence quantifying the fraction of variance attributable to the Gaussian mechanism versus the public index would help readers assess the practical utility of the private component.
  2. [Experimental results] The scalability analysis for 500 sellers and the four-benchmark comparison are welcome; adding a table that reports per-layer contribution to the final recall and latency numbers would clarify which components drive the reported operating point.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive and detailed comments. The two major concerns correctly identify that the supporting derivations for the load-bearing theoretical claims are absent from the manuscript, which weakens the significance assessment. We address each point below and will perform a major revision to supply the missing formal statements, conditions, and proofs.

read point-by-point responses

  1. Referee: [Abstract / Layer one] Abstract / Layer one: the monotone-envelope guarantee that reduces the O(Pq λ Δt) recall-loss bound looseness to 1.8–3.2× observed loss is asserted without derivation, statement of the monotonicity conditions, or proof that the envelope survives the distribution shifts the system is designed to handle; this is load-bearing for justifying the hybrid-index shortcut.

    Authors: We agree that the derivation, monotonicity conditions, and proof of survival under distribution shifts are not supplied. In the revision we will add a dedicated appendix containing (i) the precise statement of the monotonicity conditions on the neural-ODE decay, (ii) the envelope construction, and (iii) the argument that the bound remains valid when changepoints detected in Layer 2 trigger index updates. This will be placed immediately after the Layer-1 description. revision: yes

  2. Referee: [Abstract / Layer two] Abstract / Layer two: the finite-sample error guarantees for changepoint-conditioned Shapley valuation are stated to hold even after Gaussian noise is added to the affinity matrix, yet no explicit propagation of that noise through the estimator is supplied; without it the guarantee does not automatically survive the privacy mechanism.

    Authors: We agree that the manuscript states the finite-sample guarantees under noise but does not propagate the Gaussian mechanism noise through the changepoint-conditioned Shapley estimator. The revision will include an explicit noise-propagation lemma (or appendix) that composes the moments-accounting privacy noise with the estimator variance, confirming the error bounds survive the (ε,δ) guarantee. All retrieval/ranking steps remain post-processing as currently stated. revision: yes

Circularity Check

0 steps flagged

No circularity: bounds and guarantees stated without reduction to self-inputs or self-citations

full rationale

The provided abstract and description state three layers with explicit bounds (O(Pq λ Δt) recall-loss, monotone-envelope reduction to 1.8–3.2× observed loss, finite-sample Shapley error under noise, EXP3-IX regret with zCDP moments accounting) and post-processing privacy separation. No equations or text in the excerpt define any quantity in terms of itself, rename a fitted parameter as a prediction, or rely on load-bearing self-citations whose uniqueness or ansatz is imported without external verification. The monotone-envelope and noise-propagation claims are presented as results rather than shown to collapse by construction; parameters like λ and changepoint thresholds are noted as unspecified but do not trigger the enumerated circularity patterns. The derivation chain is therefore self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review provides no information on free parameters, axioms, or invented entities.

pith-pipeline@v0.9.0 · 5831 in / 1279 out tokens · 32854 ms · 2026-05-25T02:12:22.614309+00:00 · methodology

discussion (0)

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

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

79 extracted references · 79 canonical work pages · 7 internal anchors

  1. [1]

    Goodfellow, H

    Martin Abadi, Andy Chu, Ian Goodfellow, H. Brendan McMahan, Ilya Mironov, Kunal Talwar, and Li Zhang. 2016. Deep Learning with Differential Privacy. In Proceedings of the 2016 ACM SIGSAC Conference on Computer and Communications Security(Vienna, Austria)(CCS ’16). Association for Computing Machinery, New York, NY, USA, 308–318. https://doi.org/10.1145/297...

    work page doi:10.1145/2976749.2978318 2016

  2. [2]

    Ryan Prescott Adams and David J. C. MacKay. 2007. Bayesian Online Change- point Detection. arXiv:0710.3742 [stat.ML] https://arxiv.org/abs/0710.3742

    work page internal anchor Pith review Pith/arXiv arXiv 2007

  3. [3]

    Naman Agarwal and Karan Singh. 2017. The Price of Differential Privacy For Online Learning. arXiv:1701.07953 [cs.LG] https://arxiv.org/abs/1701.07953

    work page internal anchor Pith review Pith/arXiv arXiv 2017

  4. [4]

    Brendan McMahan, and Swaroop Ramaswamy

    Galen Andrew, Om Thakkar, H. Brendan McMahan, and Swaroop Ramaswamy

    work page

  5. [5]

    InProceedings of the 35th International Conference on Neural Information Processing Systems (NIPS ’21)

    Differentially private learning with adaptive clipping. InProceedings of the 35th International Conference on Neural Information Processing Systems (NIPS ’21). Curran Associates Inc., Red Hook, NY, USA, Article 1335, 12 pages

    work page

  6. [6]

    Schapire

    Peter Auer, Nicolò Cesa-Bianchi, Yoav Freund, and Robert E. Schapire

    work page

  7. [7]

    Com- put.32, 1 (2002), 48–77

    The Nonstochastic Multiarmed Bandit Problem.SIAM J. Com- put.32, 1 (2002), 48–77. https://doi.org/10.1137/S0097539701398375 arXiv:https://doi.org/10.1137/S0097539701398375

    work page doi:10.1137/s0097539701398375 2002

  8. [8]

    Santiago Andrés Azcoitia and Nikolaos Laoutaris. 2022. A Survey of Data Marketplaces and Their Business Models.SIGMOD Rec.51, 3 (Nov. 2022), 18–29. https://doi.org/10.1145/3572751.3572755

    work page doi:10.1145/3572751.3572755 2022

  9. [9]

    Emmanuel Bacry, Iacopo Mastromatteo, and Jean-François Muzy. 2015. Hawkes processes in finance. arXiv:1502.04592 [q-fin.TR] https://arxiv.org/abs/1502. 04592

    work page internal anchor Pith review Pith/arXiv arXiv 2015

  10. [10]

    Mitali Bafna and Jonathan Ullman. 2017. The Price of Selection in Differential Privacy. arXiv:1702.02970 [cs.DS] https://arxiv.org/abs/1702.02970

    work page internal anchor Pith review Pith/arXiv arXiv 2017

  11. [11]

    [n.d.].Learning from Time-Changing Data with Adaptive Windowing

    Albert Bifet and Ricard Gavaldà. [n.d.].Learning from Time-Changing Data with Adaptive Windowing. 443–448. https://doi.org/10.1137/1.9781611972771.42 arXiv:https://epubs.siam.org/doi/pdf/10.1137/1.9781611972771.42

    work page doi:10.1137/1.9781611972771.42

  12. [12]

    Vincent D Blondel, Jean-Loup Guillaume, Renaud Lambiotte, and Etienne Lefebvre. 2008. Fast unfolding of communities in large networks.Journal of Statistical Mechanics: Theory and Experiment2008, 10 (Oct. 2008), P10008. https://doi.org/10.1088/1742-5468/2008/10/p10008

    work page doi:10.1088/1742-5468/2008/10/p10008 2008

  13. [13]

    Ferdinand Brasser, Urs Müller, Alexandra Dmitrienko, Kari Kostiainen, Srdjan Capkun, and Ahmad-Reza Sadeghi. 2017. Software grand exposure: SGX cache attacks are practical. InProceedings of the 11th USENIX Conference on Offensive Technologies(Vancouver, BC, Canada)(WOOT’17). USENIX Association, USA, 11

    work page 2017

  14. [14]

    Javier Castro, Daniel Gómez, and Juan Tejada. 2009. Polynomial calculation of the Shapley value based on sampling.Comput. Oper. Res.36, 5 (May 2009), 1726–1730. https://doi.org/10.1016/j.cor.2008.04.004

    work page doi:10.1016/j.cor.2008.04.004 2009

  15. [15]

    Hubert Chan, Elaine Shi, and Dawn Song

    T.-H. Hubert Chan, Elaine Shi, and Dawn Song. 2011. Private and Continual Release of Statistics.ACM Trans. Inf. Syst. Secur.14, 3, Article 26 (Nov. 2011), 24 pages. https://doi.org/10.1145/2043621.2043626

    work page doi:10.1145/2043621.2043626 2011

  16. [16]

    Ricky T. Q. Chen, Yulia Rubanova, Jesse Bettencourt, and David Duvenaud

    work page

  17. [17]

    Neural Ordinary Differential Equations

    Neural Ordinary Differential Equations. arXiv:1806.07366 [cs.LG] https: //arxiv.org/abs/1806.07366

    work page internal anchor Pith review Pith/arXiv arXiv

  18. [18]

    Zicun Cong, Xuan Luo, Pei Jian, Feida Zhu, and Yong Zhang. 2021. Data Pricing in Machine Learning Pipelines. arXiv:2108.07915 [cs.LG] https://arxiv.org/abs/ 2108.07915

    work page arXiv 2021

  19. [19]

    Victor Costan and Srinivas Devadas. 2016. Intel SGX Explained.IACR Cryptol. ePrint Arch.2016 (2016), 86. https://api.semanticscholar.org/CorpusID:28642809

    work page 2016

  20. [20]

    Shib Sankar Dasgupta, Swayambhu Nath Ray, and Partha Talukdar. 2018. HyTE: Hyperplane-based Temporally aware Knowledge Graph Embedding. InProceed- ings of the 2018 Conference on Empirical Methods in Natural Language Process- ing(Brussels, Belgium). Association for Computational Linguistics, 2001–2011. http://aclweb.org/anthology/D18-1225

    work page 2018

  21. [21]

    Daniel Demmler, Thomas Schneider, and Michael Zohner. 2015. ABY – A Frame- work for Efficient Mixed-Protocol Secure Two-Party Computation. InProceedings of the 2015 Network and Distributed System Security Symposium (NDSS). Internet Society. https://doi.org/10.14722/ndss.2015.23113

    work page doi:10.14722/ndss.2015.23113 2015

  22. [22]

    Tim Dettmers, Pasquale Minervini, Pontus Stenetorp, and Sebastian Riedel. 2018. Convolutional 2D knowledge graph embeddings. InProceedings of the Thirty- Second AAAI Conference on Artificial Intelligence and Thirtieth Innovative Ap- plications of Artificial Intelligence Conference and Eighth AAAI Symposium on Educational Advances in Artificial Intelligenc...

    work page 2018

  23. [23]

    Dormand and P.J

    J.R. Dormand and P.J. Prince. 1980. A family of embedded Runge-Kutta formulae. J. Comput. Appl. Math.6, 1 (1980), 19–26. https://doi.org/10.1016/0771-050X(80) 90013-3

    work page doi:10.1016/0771-050x(80 1980

  24. [24]

    Jianzhang Du, Tilak Mudgal, Rutvi Rahul Gadre, Yukui Luo, and Chenghong Wang. 2024. Differentially Private Learned Indexes. arXiv:2410.21164 [cs.DB] https://arxiv.org/abs/2410.21164

    work page arXiv 2024

  25. [25]

    David Durfee and Ryan Rogers. 2019. Practical differentially private top-k selec- tion with pay-what-you-get composition. InProceedings of the 33rd International Conference on Neural Information Processing Systems. Curran Associates Inc., Red Hook, NY, USA, Article 317, 11 pages

    work page 2019

  26. [26]

    Rothblum

    Cynthia Dwork, Moni Naor, Toniann Pitassi, and Guy N. Rothblum. 2010. Differ- ential privacy under continual observation. InProceedings of the Forty-Second ACM Symposium on Theory of Computing(Cambridge, Massachusetts, USA) (STOC ’10). Association for Computing Machinery, New York, NY, USA, 715–724. https://doi.org/10.1145/1806689.1806787

    work page doi:10.1145/1806689.1806787 2010

  27. [27]

    2014.The Algorithmic Foundations of Differential Privacy

    Cynthia Dwork and Aaron Roth. 2014.The Algorithmic Foundations of Differential Privacy. Vol. 9. Now Publishers Inc., Hanover, MA, USA. 211–407 pages. https: //doi.org/10.1561/0400000042 14

    work page doi:10.1561/0400000042 2014

  28. [28]

    Vitaly Feldman and Tijana Zrnic. 2022. Individual Privacy Accounting via a Renyi Filter. arXiv:2008.11193 [cs.CR] https://arxiv.org/abs/2008.11193

    work page arXiv 2022

  29. [29]

    Franklin

    Raul Castro Fernandez, Pranav Subramaniam, and Michael J. Franklin. 2020. Data market platforms: trading data assets to solve data problems.Proc. VLDB Endow. 13, 12 (July 2020), 1933–1947. https://doi.org/10.14778/3407790.3407800

    work page doi:10.14778/3407790.3407800 2020

  30. [30]

    Kim, and James Zou

    Amirata Ghorbani, Michael P. Kim, and James Zou. 2020. A distributional framework for data valuation. InProceedings of the 37th International Conference on Machine Learning (ICML’20). JMLR.org, Article 331, 10 pages

    work page 2020

  31. [31]

    Amirata Ghorbani and James Zou. 2019. Data Shapley: Equitable Valuation of Data for Machine Learning. arXiv:1904.02868 [stat.ML] https://arxiv.org/abs/ 1904.02868

    work page internal anchor Pith review Pith/arXiv arXiv 2019

  32. [32]

    Jennifer Gillenwater, Matthew Joseph, Andrés Muñoz Medina, and Mónica Ribero. 2022. A Joint Exponential Mechanism For Differentially Private Top-𝑘. arXiv:2201.12333 [cs.CR] https://arxiv.org/abs/2201.12333

    work page arXiv 2022

  33. [33]

    Jiale Han, Austin Cheung, Yubai Wei, Zheng Yu, Xusheng Wang, Bing Zhu, and Yi Yang. 2025. RAG Meets Temporal Graphs: Time-Sensitive Modeling and Retrieval for Evolving Knowledge. arXiv:2510.13590 [cs.IR] https://arxiv.org/ abs/2510.13590

    work page arXiv 2025

  34. [34]

    Elad Hazan. 2023. Introduction to Online Convex Optimization. arXiv:1909.05207 [cs.LG] https://arxiv.org/abs/1909.05207

    work page arXiv 2023

  35. [35]

    Ruoxi Jia, David Dao, Boxin Wang, Frances Ann Hubis, Nick Hynes, Nezihe Merve Gurel, Bo Li, Ce Zhang, Dawn Song, and Costas Spanos. 2023. Towards Efficient Data Valuation Based on the Shapley Value. arXiv:1902.10275 [cs.LG] https: //arxiv.org/abs/1902.10275

    work page arXiv 2023

  36. [36]

    Woojeong Jin, Meng Qu, Xisen Jin, and Xiang Ren. 2020. Recurrent Event Network: Autoregressive Structure Inference over Temporal Knowledge Graphs. arXiv:1904.05530 [cs.LG] https://arxiv.org/abs/1904.05530

    work page arXiv 2020

  37. [37]

    Alistair E. W. Johnson, Lucas Bulgarelli, Lu Shen, Anne Gayraud, Sipanje Eraslan, Emma Rocheteau, Qinmei Huang, Jidong Cheng, Benjamin Moody, Li-wei H. Lehman, Matthew P. Lungren, Tom J. Pollard, Steven Horng, Leo Anthony Celi, and Roger G. Mark. 2023. MIMIC-IV, a freely accessible electronic health record dataset.Scientific Data10, 1 (2023), 1. https://d...

    work page doi:10.1038/s41597-022-01899-x 2023

  38. [38]

    Shiva Prasad Kasiviswanathan, Kobbi Nissim, Sofya Raskhodnikova, and Adam Smith. 2013. Analyzing graphs with node differential privacy. InProceedings of the 10th Theory of Cryptography Conference on Theory of Cryptography(Tokyo, Japan)(TCC’13). Springer-Verlag, Berlin, Heidelberg, 457–476. https://doi.org/ 10.1007/978-3-642-36594-2_26

    work page doi:10.1007/978-3-642-36594-2_26 2013

  39. [39]

    Jeremias Knoblauch, Jack Jewson, and Theodoros Damoulas. 2018. Doubly robust Bayesian inference for non-stationary streaming data with β-divergences. In Proceedings of the 32nd International Conference on Neural Information Processing Systems(Montréal, Canada)(NIPS’18). Curran Associates Inc., Red Hook, NY, USA, 64–75

    work page 2018

  40. [40]

    Paraschos Koutris, Prasang Upadhyaya, Magdalena Balazinska, Bill Howe, and Dan Suciu. 2015. Query-Based Data Pricing.J. ACM62, 5, Article 43 (Nov. 2015), 44 pages. https://doi.org/10.1145/2770870

    work page doi:10.1145/2770870 2015

  41. [41]

    Vlasis Koutsos, Dimitrios Papadopoulos, Dimitris Chatzopoulos, Sasu Tarkoma, and Pan Hui. 2020. Agora: A Privacy-aware Data Marketplace. In2020 IEEE 40th International Conference on Distributed Computing Systems (ICDCS). 1211–1212. https://doi.org/10.1109/ICDCS47774.2020.00156

    work page doi:10.1109/icdcs47774.2020.00156 2020

  42. [42]

    Yongchan Kwon and James Zou. 2022. Beta Shapley: a Unified and Noise-reduced Data Valuation Framework for Machine Learning. arXiv:2110.14049 [cs.LG] https://arxiv.org/abs/2110.14049

    work page arXiv 2022

  43. [43]

    Julien Leblay and Melisachew Wudage Chekol. 2018. Deriving Validity Time in Knowledge Graph. InCompanion Proceedings of the The Web Conference 2018 (Lyon, France)(WWW ’18). International World Wide Web Conferences Steering Committee, Republic and Canton of Geneva, CHE, 1771–1776. https://doi.org/ 10.1145/3184558.3191639

    work page doi:10.1145/3184558.3191639 2018

  44. [44]

    Yehuda Lindell. 2021. Fast Secure Two-Party ECDSA Signing.J. Cryptol.34, 4 (Oct. 2021), 38. https://doi.org/10.1007/s00145-021-09409-9

    work page doi:10.1007/s00145-021-09409-9 2021

  45. [45]

    Jinfei Liu, Jian Lou, Junxu Liu, Li Xiong, Jian Pei, and Jimeng Sun. 2021. Dealer: an end-to-end model marketplace with differential privacy.Proc. VLDB Endow. 14, 6 (Feb. 2021), 957–969. https://doi.org/10.14778/3447689.3447700

    work page doi:10.14778/3447689.3447700 2021

  46. [46]

    Shige Liu, Zhifang Zeng, Li Chen, Adil Ainihaer, Arun Ramasami, Songting Chen, Yu Xu, Mingxi Wu, and Jianguo Wang. 2025. TigerVector: Supporting Vector Search in Graph Databases for Advanced RAGs. InCompanion of the 2025 International Conference on Management of Data(Berlin, Germany)(SIG- MOD/PODS ’25). Association for Computing Machinery, New York, NY, U...

    work page doi:10.1145/3722212.3724456 2025

  47. [47]

    Ryan Lowe, Yi Wu, Aviv Tamar, Jean Harb, Pieter Abbeel, and Igor Mordatch

    work page

  48. [48]

    InProceedings of the 31st International Conference on Neural Information Processing Systems(Long Beach, California, USA)(NIPS’17)

    Multi-agent actor-critic for mixed cooperative-competitive environments. InProceedings of the 31st International Conference on Neural Information Processing Systems(Long Beach, California, USA)(NIPS’17). Curran Associates Inc., Red Hook, NY, USA, 6382–6393

    work page

  49. [49]

    Malkov and D

    Yu A. Malkov and D. A. Yashunin. 2020. Efficient and Robust Approximate Nearest Neighbor Search Using Hierarchical Navigable Small World Graphs. IEEE Trans. Pattern Anal. Mach. Intell.42, 4 (April 2020), 824–836. https://doi. org/10.1109/TPAMI.2018.2889473

    work page doi:10.1109/tpami.2018.2889473 2020

  50. [50]

    Ryan McKenna and Daniel Sheldon. 2020. Permute-and-flip: a new mechanism for differentially private selection. InProceedings of the 34th International Conference on Neural Information Processing Systems(Vancouver, BC, Canada)(NIPS ’20). Curran Associates Inc., Red Hook, NY, USA, Article 17, 11 pages

    work page 2020

  51. [51]

    Frank McSherry and Kunal Talwar. 2007. Mechanism Design via Differential Privacy. InProceedings of the 48th Annual IEEE Symposium on Foundations of Computer Science (FOCS ’07). IEEE Computer Society, USA, 94–103. https: //doi.org/10.1109/FOCS.2007.41

    work page doi:10.1109/focs.2007.41 2007

  52. [52]

    Ilya Mironov. 2017. Rényi Differential Privacy. In2017 IEEE 30th Computer Security Foundations Symposium (CSF). IEEE, 263–275. https://doi.org/10.1109/ csf.2017.11

    work page 2017

  53. [53]

    Ilyas, Theodoros Rekatsinas, and Shivaram Venkataraman

    Jason Mohoney, Devesh Sarda, Mengze Tang, Shihabur Rahman Chowdhury, Anil Pacaci, Ihab F. Ilyas, Theodoros Rekatsinas, and Shivaram Venkataraman

    work page

  54. [54]

    , Article 9 (2025), 17 pages

    Quake: adaptive indexing for vector search. , Article 9 (2025), 17 pages

    work page 2025

  55. [55]

    Pranay Mundra, Charalampos Papamanthou, Julian Shun, and Quanquan C. Liu

    work page

  56. [56]

    arXiv:2506.20828 [cs.DS] https://arxiv.org/abs/2506.20828

    Practical and Accurate Local Edge Differentially Private Graph Algorithms. arXiv:2506.20828 [cs.DS] https://arxiv.org/abs/2506.20828

    work page arXiv

  57. [58]

    Gergely Neu. 2015. Explore no more: improved high-probability regret bounds for non-stochastic bandits. InProceedings of the 29th International Conference on Neural Information Processing Systems - Volume 2(Montreal, Canada)(NIPS’15). MIT Press, Cambridge, MA, USA, 3168–3176

    work page 2015

  58. [59]

    Kobbi Nissim, Sofya Raskhodnikova, and Adam Smith. 2007. Smooth sensitivity and sampling in private data analysis. InProceedings of the Thirty-Ninth Annual ACM Symposium on Theory of Computing(San Diego, California, USA)(STOC ’07). Association for Computing Machinery, New York, NY, USA, 75–84. https: //doi.org/10.1145/1250790.1250803

    work page doi:10.1145/1250790.1250803 2007

  59. [60]

    E. S. PAGE. 1954. CONTINUOUS INSPECTION SCHEMES.Biometrika 41, 1-2 (06 1954), 100–115. https://doi.org/10.1093/biomet/41.1-2.100 arXiv:https://academic.oup.com/biomet/article-pdf/41/1-2/100/1243987/41-1-2- 100.pdf

    work page doi:10.1093/biomet/41.1-2.100 1954

  60. [62]

    Papadimitriou and John N

    Christos H. Papadimitriou and John N. Tsitsiklis. 1987. The Complexity of Markov Decision Processes.Math. Oper. Res.12, 3 (Aug. 1987), 441–450

    work page 1987

  61. [63]

    Liana Patel, Peter Kraft, Carlos Guestrin, and Matei Zaharia. 2024. ACORN: Performant and Predicate-Agnostic Search Over Vector Embeddings and Struc- tured Data.Proc. ACM Manag. Data2, 3, Article 120 (May 2024), 27 pages. https://doi.org/10.1145/3654923

    work page doi:10.1145/3654923 2024

  62. [64]

    Ryan Rogers, Aaron Roth, Jonathan Ullman, and Salil Vadhan. 2016. Privacy odometers and filters: pay-as-you-go composition. InProceedings of the 30th International Conference on Neural Information Processing Systems(Barcelona, Spain)(NIPS’16). Curran Associates Inc., Red Hook, NY, USA, 1929–1937

    work page 2016

  63. [65]

    Gaurav Sehgal and Semih Salihoglu. 2025. NaviX: A Native Vector Index De- sign for Graph DBMSs With Robust Predicate-Agnostic Search Performance. arXiv:2506.23397 [cs.IR] https://arxiv.org/abs/2506.23397

    work page arXiv 2025

  64. [66]

    Lloyd S. Shapley. 1953.A Value for n-Person Games. Princeton University Press, Princeton, NJ. 307–318 pages. https://doi.org/10.1515/9781400881970-018

    work page doi:10.1515/9781400881970-018 1953

  65. [67]

    Aditi Singh, Suhas Jayaram Subramanya, Ravishankar Krishnaswamy, and Har- sha Vardhan Simhadri. 2021. FreshDiskANN: A Fast and Accurate Graph- Based ANN Index for Streaming Similarity Search. arXiv:2105.09613 [cs.IR] https://arxiv.org/abs/2105.09613

    work page arXiv 2021

  66. [68]

    Aristide Tossou and Christos Dimitrakakis. 2015. Algorithms for Differentially Private Multi-Armed Bandits. arXiv:1511.08681 [stat.ML] https://arxiv.org/abs/ 1511.08681

    work page internal anchor Pith review Pith/arXiv arXiv 2015

  67. [69]

    Wang and Ruoxi Jia

    Jiachen T. Wang and Ruoxi Jia. 2023. Data Banzhaf: A Robust Data Valuation Framework for Machine Learning. arXiv:2205.15466 [cs.LG] https://arxiv.org/ abs/2205.15466

    work page arXiv 2023

  68. [70]

    Hao Wu and Hanwen Zhang. 2024. Faster differentially private top-k selection: a joint exponential mechanism with pruning. InProceedings of the 38th Inter- national Conference on Neural Information Processing Systems(Vancouver, BC, Canada)(NIPS ’24). Curran Associates Inc., Red Hook, NY, USA, Article 2266, 27 pages

    work page 2024

  69. [71]

    Mengmeng Wu, Ruoxi Jia, Changle Lin, Wei Huang, and Xiangyu Chang. 2023. Variance reduced Shapley value estimation for trustworthy data valuation.Com- put. Oper. Res.159, C (Nov. 2023), 9. https://doi.org/10.1016/j.cor.2023.106305

    work page doi:10.1016/j.cor.2023.106305 2023

  70. [72]

    Yulian Wu, Xingyu Zhou, Youming Tao, and Di Wang. 2023. On private and robust bandits. InProceedings of the 37th International Conference on Neural Information Processing Systems(New Orleans, LA, USA)(NIPS ’23). Curran Associates Inc., Red Hook, NY, USA, Article 1511, 13 pages

    work page 2023

  71. [73]

    Wentao Xiao, Yueyang Zhan, Rui Xi, Mengshu Hou, and Jianming Liao. 2024. Enhancing HNSW Index for Real-Time Updates: Addressing Unreachable Points and Performance Degradation. arXiv:2407.07871 [cs.IR] https://arxiv.org/abs/ 2407.07871

    work page arXiv 2024

  72. [74]

    Da Xu, Chuanwei Ruan, Evren Korpeoglu, Sushant Kumar, and Kannan Achan. 2020. Inductive Representation Learning on Temporal Graphs. 15 arXiv:2002.07962 [cs.LG] https://arxiv.org/abs/2002.07962

    work page arXiv 2020

  73. [75]

    Yuming Xu, Hengyu Liang, Jin Li, Shuotao Xu, Qi Chen, Qianxi Zhang, Cheng Li, Ziyue Yang, Fan Yang, Yuqing Yang, Peng Cheng, and Mao Yang. 2023. SPFresh: Incremental In-Place Update for Billion-Scale Vector Search. InProceedings of the 29th Symposium on Operating Systems Principles(Koblenz, Germany)(SOSP ’23). Association for Computing Machinery, New York...

    work page doi:10.1145/3600006.3613166 2023

  74. [76]

    Jinfa Yang, Xianghua Ying, Yongjie Shi, and Bowei Xing. 2024. Tensor decompo- sitions for temporal knowledge graph completion with time perspective•.Expert Syst. Appl.237, PA (March 2024), 12. https://doi.org/10.1016/j.eswa.2023.121267

    work page doi:10.1016/j.eswa.2023.121267 2024

  75. [77]

    Yelp. 2026. Yelp Open Dataset. https://www.yelp.com/dataset

    work page 2026

  76. [78]

    Quan Yuan, Zhikun Zhang, Linkang Du, Min Chen, Mingyang Sun, Yunjun Gao, Michael Backes, Shibo He, and Jiming Chen. 2025. PSGraph: Differen- tially Private Streaming Graph Synthesis by Considering Temporal Dynamics. arXiv:2412.11369 [cs.CR] https://arxiv.org/abs/2412.11369

    work page arXiv 2025

  77. [79]

    Qiuchen Zhang, Hong kyu Lee, Jing Ma, Jian Lou, Carl Yang, and Li Xiong. 2024. DPAR: Decoupled Graph Neural Networks with Node-Level Differential Privacy. InProceedings of the ACM Web Conference 2024(Singapore, Singapore)(WWW ’24). Association for Computing Machinery, New York, NY, USA, 1170–1181. https://doi.org/10.1145/3589334.3645531

    work page doi:10.1145/3589334.3645531 2024

  78. [80]

    Ziyu Zhang, Yuanhao Wei, Joshua Engels, and Julian Shun. 2025. CleANN: Efficient Full Dynamism in Graph-based Approximate Nearest Neighbor Search. arXiv:2507.19802 [cs.DB] https://arxiv.org/abs/2507.19802

    work page arXiv 2025

  79. [81]

    Xiaoyao Zhong, Haotian Li, Jiabao Jin, Mingyu Yang, Deming Chu, Xiangyu Wang, Zhitao Shen, Wei Jia, George Gu, Yi Xie, Xuemin Lin, Heng Tao Shen, Jingkuan Song, and Peng Cheng. 2025. VSAG: An Optimized Search Framework for Graph-Based Approximate Nearest Neighbor Search. , 14 pages. https: //doi.org/10.14778/3750601.3750624 16

    work page doi:10.14778/3750601.3750624 2025