Pith. sign in

REVIEW 4 major objections 5 minor 91 references

Stream DaQ: Stream-First Data Quality Monitoring

T0 review · 4 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read Window-based, stream-first data quality monitoring catches temporally localized errors that static and incremental tools miss, and its implementation runs up to 13.8x faster on small windows.

desk verdict A useful streaming data quality framework with an open-source implementation, but the evaluation overstates performance due to asymmetric timing and never tests whether the checks actually detect errors. read the letter →

arxiv 2506.06147 v1 pith:BAKCVIVT submitted 2025-06-06 cs.DB

Vasileios Papastergios , Anastasios Gounaris This is my paper · ORCID
classification cs.DB
keywords dataqualitymonitoringstreamswindowingmeta-streamstreamprocessingPythonframeworkreal-time
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper tries to establish that data quality monitoring for unbounded streams must be stream-first: checks should run continuously over configurable temporal windows, not as one-off static assessments or as incremental aggregates accumulated since stream inception. It formalizes a model with windowed measurement, dynamic constraint adaptation, and a quality meta-stream, and presents Stream DaQ, an open-source Python implementation that unifies more than 30 checks from static tools. The evaluation claims that Stream DaQ significantly outperforms a production-grade alternative in execution time and throughput, up to 13.8x for 1-minute tumbling windows, while offering richer native streaming functionality. A sympathetic reader would care because timely, context-aware detection of temporally localized errors can prevent bad data from propagating through real-time pipelines and downstream ML models.

What carries the argument

The central object is the quality meta-stream: a structured output stream whose entries are tuples of window start time, window end time, a measurement value, and an assessment result, produced by continuously applying stream quality assessments over a sequence of windows. It carries the argument because it turns monitoring from a one-off verdict into a live, downstream-consumable signal, and it is generated by the model's configurable windowing, dynamic constraint adaptation, and compositional check categories (tuple-at-a-time, window context, reference data, dynamically adapted context, and keyed checks).

What would settle it

Inject known errors of controlled duration into a stream, such as five-minute bursts of NULLs, duplicates, or placeholder values, and measure whether windowed checks raise alerts with high precision and recall within the expected window; if the checks miss short-lived errors or fire false alarms on clean windows, the central claim collapses.

Watch

Extended reading notes

Core claim

The paper's central claim is that data quality monitoring for unbounded streams should be modeled as a continuous streaming computation: assessments are applied over a sequence of user-configurable windows, constraints can adapt using recent context rather than stream-inception aggregates, and the results flow onward as a quality meta-stream. It argues that this design detects temporally dependent errors, such as bursts of missing values, duplicates, or placeholder fares, that static nightly checks detect too late and incremental running-aggregate checks mask or report after the bad period has ended. The paper further claims that the implementation, Stream DaQ, delivers substantial performance benefits in exactly the real-time scenarios that matter: for small windows it outperforms a production-grade batch-extended baseline by up to 13.8x, and it remains competitive for larger windows while providing a broader, compositionally expressive check suite.

Load-bearing premise

The practical usefulness of the model rests on the unvalidated assumption that quality checks originally designed for static, whole-dataset analysis remain accurate and meaningful when run over short windows of an unbounded stream; the evaluation measures execution time, throughput, and feature coverage, but never injects known errors to test whether the windowed alerts actually detect them.

Editorial extensions

If this is right

  • Short-window checks can detect bursts of errors, such as 20% missing destinations in five minutes, that a running average since stream inception would mask or report only after the low-quality period has ended.
  • Quality meta-streams can be consumed by downstream systems as real-time awareness signals, enabling alerts, routing of erroneous elements to separate streams, or pausing model updates during detected low-quality intervals.
  • Composing check categories yields entity-specific, context-aware constraints, such as per-taxi fare deviations against a rolling two-hour baseline, rather than generic whole-dataset rules.
  • Small tumbling and moderately overlapping sliding windows strongly favor the lightweight Python-native design, with improvements up to 13.8x; larger windows and 90% overlap configurations make performance comparable to the batch-oriented distributed baseline.
  • The Python-native interface lowers the barrier to integrating continuous quality monitoring into existing data science workflows, including pandas-based preprocessing and Jupyter-based analysis.

Reading between the lines

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

  • If detection quality is confirmed through error-injection experiments, the same window-based template could transfer to other domains where temporal locality matters, such as fraud detection, sensor validation, and network monitoring.
  • The comparable performance at 90% sliding-window overlap suggests an incremental sliding-window optimization opportunity; the authors themselves note that incremental processing is a direction for future work.
  • The quality meta-stream could plausibly be extended to serve as a gating signal for online learning, telling a model to skip weight updates or to down-weight inputs during windows flagged as low-quality.
  • A direct comparison of false-positive rates between windowed checks and static checks on the same stream would test whether the expressive model actually reduces misleading alerts, not just detection latency.
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

4 major / 5 minor

Summary. The paper proposes a stream-first data quality monitoring model for unbounded data streams. The model is built around configurable time windows, dynamic constraint adaptation, and a continuous 'quality meta-stream' output, formalized in Definitions 3.1 through 3.4. The authors also present Stream DaQ, an open-source Python framework on Pathway that implements the model and unifies more than 30 quality checks adapted from static tools. The evaluation compares Stream DaQ with a Spark-based Deequ adaptation on replayed Reddit comment data, reporting throughput, latency, and per-window execution time, with claimed speedups up to 13.8x for small windows, plus a qualitative functional comparison with Apache Griffin and Deequ.

Significance. If the performance and utility claims are substantiated, the paper addresses a real gap: most data quality tooling assumes static or incrementally growing data, while streaming pipelines need temporally localized, continuous quality assessment. The formal model is clearly presented, the compositional check categories are expressive, and the open-source implementation with public experiment code is a concrete strength. However, the current evidence is not yet sufficient: the main quantitative comparison uses an asymmetric cost boundary, the results lack any measure of variance, and the central promise of timely, context-aware detection is never tested by injecting known errors. These are fixable within the manuscript's scope, so the work is promising but needs a major revision.

major comments (4)
  1. [Section 5.1 (Experimental Setup)] The measurement of 'net processing time' is asymmetric: the text states 'For Deequ, we measure net processing time excluding Spark's windowing, while for Stream DaQ the net processing time stands for both windowing and execution of quality checks.' This directly biases the per-window execution-time comparisons and the headline 13.8x speedup in favor of Stream DaQ, since windowing is part of the streaming pipeline for both systems. Please redo the comparison with identical cost boundaries (either include windowing for both or exclude it for both) and report the windowing and check-execution components separately.
  2. [Section 5.1 (Results)] All reported numbers are averages over five or more runs, but no standard deviations, confidence intervals, or per-run distributions are given, and Figures 4 and 5 show no error bars. The claims of 'significant' performance advantages for small windows and 'comparable' performance for large windows cannot be assessed without a measure of variability. Please report variance or distributional information, and state whether the observed differences are stable across runs and resource configurations.
  3. [Section 3.1 and Section 5] The paper's central practical claim is 'timely, temporally fine-grained detection' of generic and temporally dependent errors, but Section 5 evaluates only execution time, throughput, latency, and feature coverage. No experiment injects known quality errors into the stream to measure detection rate, false-alarm rate, or time-to-detection. This is load-bearing: a system can process windows quickly while still missing the temporally localized errors it is designed to catch. Please add detection-effectiveness experiments, e.g., injecting missing values, placeholder bursts, duplicates, and out-of-order events, and reporting precision/recall and detection delay for Stream DaQ and the adapted Deequ baseline.
  4. [Section 5.1 (Evaluation Strategy and Setup)] The comparison workload is not specified precisely. The text says the evaluation 'utilizes only a subset of Stream DaQ's model expressiveness that corresponds to functionality available in Deequ's static data capabilities,' but it does not enumerate which checks were used, how many constraints were active, what thresholds were set, or how the checks were mapped between the two systems. Without this information, the throughput and execution-time results cannot be reproduced or interpreted as a meaningful comparison. Please provide the concrete check configuration used in the experiments.
minor comments (5)
  1. [Section 2.2.2] There is a typo in 'also knwon as dimensions' — should be 'known'.
  2. [References] References [6] and [7] are duplicate entries for the same Akidau et al. paper on the dataflow model; one of them should be removed or replaced with the intended distinct citation.
  3. [Section 5.1 (Main Setup Details)] The text gives five average numbers of records per window (1783, 2276, 5731, 20035, and 45107), but the experimental matrix includes ten window configurations (four tumbling and six sliding); please clarify which configurations these averages correspond to.
  4. [Table 2] The formatting of Table 2 appears garbled, with stray 'z}|{' artifacts in the column head area; the table should be regenerated so that the columns align correctly.
  5. [Figures 4 and 5] The axis labels and legends in Figures 4 and 5 are very small, and some y-axis units (e.g., '1e-9 sec' and '0e-4 sec') are hard to read; larger fonts and clearer unit labels would improve interpretability.

Circularity Check

0 steps flagged · score 1.0 of 10

No circular derivation: windowed checks are defined directly, performance is externally benchmarked against Deequ, and the sole self-citation (the authors' survey) is not load-bearing.

full rationale

The paper's formal content is definitional rather than derived-from-data: Definitions 2.3 through 3.4 introduce stream windows, stream quality measurement, assessment, monitoring, and the quality meta-stream as explicit constructions. No quantity is fitted to a subset of data and then reported as a prediction, and no equation is shown to be equivalent to its own input by construction. The performance claims in Section 5 are grounded in an external baseline, Amazon's Deequ, measured on the Reddit Comments dataset; throughput, latency, and execution time are directly measured rather than inferred from the model's assumptions. The Deequ streaming wrapper was written by the same team, which is a benchmarking-fairness risk, but it does not make the measured comparison circular because the baseline implementation is an independently developed production system. The only self-citation that appears in a load-bearing position is reference [51], the authors' own survey, used in Section 4.2 to justify the inventory of 'over 30 fragmented quality checks' drawn from seven named static tools. That citation is not a uniqueness theorem, does not forbid alternatives, and the underlying tools and checks are transparently listed in Table 1, making the claim externally verifiable; it is therefore a minor self-citation rather than a circular step. The paper does not empirically validate detection quality by injecting known errors, so the practical utility claim rests on an untested assumption about false alarms and temporally localized error detection, but that is an evidentiary gap, not circularity: the windowed checks would still behave exactly as defined even if the evaluation were expanded. Overall, the central contributions survive as independent content, and no circular reduction was found.

Assumptions & free parameters 0 free parameters · 4 assumptions · 1 invented entities

The theoretical model has no fitted constants; window durations and thresholds in the evaluation are user choices, not fitted parameters. The load-bearing assumptions are standard stream-processing premises (reliable event-time timestamps, correct windowing) plus the unvalidated transfer of static checks to windows. The Deequ baseline fairness is an ad hoc assumption introduced by the evaluation design.

assumptions (4)
  • domain assumption Event-time timestamps are present and reliable for every stream element.
    Definition 2.1 requires a creation timestamp, and Definition 2.3 defines windows over event-time timestamps. Late-arrival handling in Section 4.1 also depends on trusting event time.
  • domain assumption Pathway's engine correctly implements the window semantics and reducers used by Stream DaQ.
    Section 4.1 and Section 7.3 delegate windowing and reduction to Pathway's Rust backend. No formal verification of these semantics is provided.
  • domain assumption Static quality checks remain semantically meaningful when restricted to a finite window of an unbounded stream.
    Section 3.3 and Section 4.2 adapt checks from static tools to windows, but no detection-accuracy experiment validates this transfer.
  • ad hoc to paper The benchmark adaptation of Deequ via Spark Streaming is a fair representation of Deequ's capabilities.
    Section 5.1 describes a self-authored wrapper around Deequ. Its fidelity to production-grade Deequ is not independently validated, and the timing methodology excludes Spark windowing for Deequ but not for Stream DaQ.
invented entities (1)
  • Quality meta-stream
    purpose: A continuous sequence of window measurement-assessment tuples consumed by downstream systems as a quality-awareness signal.
    Defined in Definition 3.4 and exposed through daq.watch_out() in Listing 1. It is observable, but no external evaluation demonstrates its downstream utility.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Stream DaQ: Stream-First Data Quality Monitoring." pith.science (2026). https://pith.science/paper/BAKCVIVT

@misc{pith2026250606147,
  author       = {Pith},
  title        = {Pith review of: Stream DaQ: Stream-First Data Quality Monitoring},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BAKCVIVT}},
  note         = {Machine review of arXiv:2506.06147}
}
read the original abstract

Data quality is fundamental to modern data science workflows, where data continuously flows as unbounded streams feeding critical downstream tasks, from elementary analytics to advanced artificial intelligence models. Existing data quality approaches either focus exclusively on static data or treat streaming as an extension of batch processing, lacking the temporal granularity and contextual awareness required for true streaming applications. In this paper, we present a novel data quality monitoring model specifically designed for unbounded data streams. Our model introduces stream-first concepts, such as configurable windowing mechanisms, dynamic constraint adaptation, and continuous assessment that produces quality meta-streams for real-time pipeline awareness. To demonstrate practical applicability, we developed Stream DaQ, an open-source Python framework that implements our theoretical model. Stream DaQ unifies and adapts over 30 quality checks fragmented across existing static tools into a comprehensive streaming suite, enabling practitioners to define sophisticated, context-aware quality constraints through compositional expressiveness. Our evaluation demonstrates that the model's implementation significantly outperforms a production-grade alternative in both execution time and throughput while offering richer functionality via native streaming capabilities compared to other choices. Through its Python-native design, Stream DaQ seamlessly integrates with modern data science workflows, making continuous quality monitoring accessible to the broader data science community.

Figures

Figures reproduced from arXiv: 2506.06147 by the authors.

Figure 1
Figure 1. Real-time errors in NYC taxi data [38] can be de￾tected late, falsely, or completely missed when applying static (top) or incremental (bottom) data quality checks. Limitations of Previous Approaches. Although well-established, current quality assessment methodologies [9, 10, 82] and implemen￾tations [22, 71] address a different problem. They adopt either static or incremental [68] data models, but not true streaming… view at source ↗
Figure 2
Figure 2. A real-world motivating example from the NYC taxi dataset [ [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. The Stream DaQ framework architecture [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Comparison between Stream DaQ and Amazon’s Deequ [69] across varying window settings, number of input records and resource constraints on single machine for the Reddit comments dataset. in pure Python, typically more expressive and familiar than Spark SQL for modern da…
Figure 5
Figure 5. Figure 5: Latency (left) and throughput (right) comparison be [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

91 extracted references · 57 canonical work pages

  1. [1]

    [n. d.]. Kafka Streams Documentation. https://kafka.apache.org/documentation/ streams/. Accessed: 2025-01-20

    2025

  2. [2]

    Daniel Abadi, Rakesh Agrawal, Anastasia Ailamaki, Magdalena Balazinska, Philip A Bernstein, Michael J Carey, Surajit Chaudhuri, Jeffrey Dean, AnHai Doan, Michael J Franklin, et al. 2016. The beckman report on database research. Commun. ACM 59, 2 (2016), 92–99

    2016

  3. [3]

    Mohamed Abdelaal, Christian Hammacher, and Harald Schöning. 2023. REIN: A Comprehensive Benchmark Framework for Data Cleaning Methods in ML Pipelines. https://doi.org/10.48786/EDBT.2023.43

    doi:10.48786/edbt.2023.43 2023

  4. [4]

    Ziawasch Abedjan, Lukasz Golab, and Felix Naumann. 2015. Profiling relational data: a survey. The VLDB Journal 24, 4 (Aug. 2015), 557–581. https://doi.org/10. 1007/s00778-015-0389-y

    2015

  5. [5]

    Rakesh Agrawal, Anastasia Ailamaki, Philip A Bernstein, Eric A Brewer, Michael J Carey, Surajit Chaudhuri, AnHai Doan, Daniela Florescu, Michael J Franklin, Hector Garcia-Molina, et al. 2008. The Claremont report on database research. ACM SIGMOD Record 37, 3 (2008), 9–19

    2008

  6. [7]

    Fernández-Moctezuma, Reuven Lax, Sam McVeety, Daniel Mills, Frances Perry, Eric Schmidt, and Sam Whittle

    Tyler Akidau, Robert Bradshaw, Craig Chambers, Slava Chernyak, Rafael J. Fernández-Moctezuma, Reuven Lax, Sam McVeety, Daniel Mills, Frances Perry, Eric Schmidt, and Sam Whittle. 2015. The dataflow model: a practical approach to balancing correctness, latency, and cost in massive-scale, unbounded, out-of- order data processing. Proc. VLDB Endow. 8, 12 (Au...

    arXiv 2015

  7. [8]

    Michal Bartoszkiewicz, Jan Chorowski, Adrian Kosowski, Jakub Kowalski, Sergey Kulik, Mateusz Lewandowski, Krzysztof Nowicki, Kamil Piechowiak, Olivier Ruas, Zuzanna Stamirowska, and Przemyslaw Uznanski. 2023. Pathway: a fast and flexible unified stream data processing framework for analytical and Machine Learning applications. arXiv:2307.13116 [cs.LG] htt...

    work page Pith review arXiv 2023

  8. [9]

    Carlo Batini, Cinzia Cappiello, Chiara Francalanci, and Andrea Maurino. 2009. Methodologies for Data Quality Assessment and Improvement. Comput. Surveys 41, 3 (2009), 1–52

    2009

Show all 91 references
  1. [10]

    Carlo Batini and Monica Scannapieco. 2016. Data and Information Quality: Dimensions, Principles and Techniques. Springer International Publishing. https: //doi.org/10.1007/978-3-319-24106-7

    2016 doi

  2. [11]

    Eric Breck, Marty Zinkevich, Neoklis Polyzotis, Steven Whang, and Sudip Roy

  3. [12]

    Hong Cai, Boyi Xu, Lihong Jiang, and Athanasios V Vasilakos. 2017. IoT-based big data storage systems in cloud computing: Perspectives and challenges. IEEE Internet of Things Journal 4, 1 (2017), 75–87

    2017

  4. [13]

    Paris Carbone, Asterios Katsifodimos, Stephan Ewen, Volker Markl, Seif Haridi, and Kostas Tzoumas. 2015. Apache Flink™: Stream and Batch Processing in a Single Engine. IEEE Data Engineering Bulletin 38 (01 2015)

    2015

  5. [14]

    Oscar Castro, Pierrick Bruneau, Jean-Sébastien Sottet, and Dario Torregrossa

  6. [15]

    Ilyas, Mourad Ouzzani, Paolo Papotti, Nan Tang, and Yin Ye

    Xu Chu, John Morcos, Ihab F. Ilyas, Mourad Ouzzani, Paolo Papotti, Nan Tang, and Yin Ye. 2015. KATARA: A Data Cleaning System Powered by Knowledge Bases and Crowdsourcing. In Proceedings of the 2015 ACM SIGMOD International Conference on Management of Data (Melbourne, Victoria...

    2015

  7. [16]

    Ilyas, Mourad Ouzzani, and Nan Tang

    Michele Dallachiesa, Amr Ebaid, Ahmed Eldawy, Ahmed Elmagarmid, Ihab F. Ilyas, Mourad Ouzzani, and Nan Tang. 2013. NADEEF: a commodity data cleaning system. In Proceedings of the 2013 ACM SIGMOD International Conference on Management of Data (New York, New York, USA) (SIGMOD ’...

    2013

  8. [17]

    dbt Core GitHub Repository. 2025. dbt Core GitHub Repository. https://github. com/dbt-labs/dbt-core

    2025

  9. [18]

    Jeremy Debattista, SÖren Auer, and Christoph Lange. 2016. Luzzu—A Methodol- ogy and Framework for Linked Data Quality Assessment.J. Data and Information Quality 8, 1, Article 4 (Oct. 2016), 32 pages. https://doi.org/10.1145/2992786

    2016 doi

  10. [19]

    Marcos Dias de Assunção, Alexandre da Silva Veith, and Rajkumar Buyya. 2018. Distributed data stream processing and edge computing: A survey on resource elasticity and future directions. Journal of Network and Computer Applications 103 (2018), 1–17. https://doi.org/10.1016/j.j...

    2018 doi

  11. [20]

    Alberto Diez-Olivan, Javier Del Ser, Diego Galar, and Basilio Sierra. 2019. Data fusion and machine learning for industrial prognosis: Trends and perspectives towards Industry 4.0. Information Fusion 50 (2019), 92–111. https://doi.org/10. 1016/j.inffus.2018.10.005

    2019

  12. [21]

    Lisa Ehrlinger, Bernhard Werth, and Wolfram Wöß. 2023. Automating Data Quality Monitoring with Reference Data Profiles . Springer Nature Switzerland, 24–44. https://doi.org/10.1007/978-3-031-37890-4_2

    2023 doi

  13. [22]

    Lisa Ehrlinger and Wolfram Wöß. 2022. A Survey of Data Quality Measurement and Monitoring Tools. Frontiers in Big Data 5 (2022). https://doi.org/10.3389/ fdata.2022.850611

    2022

  14. [23]

    Ailamaki et al

    A. Ailamaki et al. 2025. The Cambridge Report on Database Research. https: //doi.org/10.48550/ARXIV.2504.11259

    2025 doi arXiv

  15. [24]

    Sedir Mohammed et al. 2025. The effects of data quality on machine learning performance on tabular data. Information Systems 132 (July 2025), 102549. https: //doi.org/10.1016/j.is.2025.102549

    2025

  16. [25]

    Hadi Fadlallah, Rima Kilany, Houssein Dhayne, Rami El Haddad, Rafiqul Haque, Yehia Taher, and Ali Jaber. 2023. Context-aware Big Data Quality Assessment: A Scoping Review. J. Data and Information Quality 15, 3, Article 25 (Aug. 2023), 33 pages. https://doi.org/10.1145/3603707

    2023 doi

  17. [26]

    Donatella Firmani, Barna Saha, and Divesh Srivastava. 2016. Online entity resolution using an Oracle. 9, 5 (Jan. 2016), 384–395. https://doi.org/10.14778/ 2876473.2876474

    2016

  18. [27]

    International Organization for Standardization. 2008. ISO/IEC 25012: Software Engineering: Software Product Quality Requirements and Evaluation (SQuaRE): Data Quality Model . ISO/IEC. https://iso25000.com/index.php/en/iso-25000- standards/iso-25012

    2008

  19. [28]

    Apache Software Foundation. [n. d.]. Apache IoTDB Official Website. https: //iotdb.apache.org/. Accessed: 2025-02-20

    2025

  20. [29]

    Marios Fragkoulis, Paris Carbone, Vasiliki Kalavri, and Asterios Katsifodimos

  21. [30]

    Mohamed Medhat Gaber, Arkady Zaslavsky, and Shonali Krishnaswamy. 2005. Mining data streams: a review. SIGMOD Rec. 34, 2 (June 2005), 18–26. https: //doi.org/10.1145/1083784.1083789

    2005

  22. [31]

    João Gama, Indrundefined Žliobaitundefined, Albert Bifet, Mykola Pechenizkiy, and Abdelhamid Bouchachia. 2014. A survey on concept drift adaptation. ACM Comput. Surv. 46, 4, Article 44 (March 2014), 37 pages. https://doi.org/10.1145/ 2523813

    2014

  23. [32]

    2023), 507–541

    A survey on the evolution of stream processing systems.The VLDB Journal 33, 2 (Nov. 2023), 507–541. https://doi.org/10.1007/s00778-023-00819-8

    2023 doi

  24. [33]

    Tamer Özsu

    Lukasz Golab and M. Tamer Özsu. 2010. Data Stream Management . Springer International Publishing. https://doi.org/10.1007/978-3-031-01837-4

    2010 doi

  25. [34]

    Rong Gu, Yang Qi, Tongyu Wu, Zhaokang Wang, Xiaolong Xu, Chunfeng Yuan, and Yihua Huang. 2021. SparkDQ: Efficient generic big data quality management on distributed data-parallel computation. J. Parallel and Distrib. Comput. 156 (2021), 35–49

    2021

  26. [35]

    Gassman, Walter W

    Jennifer J. Gassman, Walter W. Owen, Timothy E. Kuntz, Jeffrey P. Martin, and William P. Amoroso. 1995. Data quality assurance, monitoring, and reporting. Controlled Clinical Trials 16, 2 (April 1995), 104–136. https://doi.org/10.1016/ 0197-2456(94)00095-k

    1995

  27. [36]

    TIOBE Programming Community Index. [n. d.]. Apache Griffin’s Official Docker Guide. https://www.tiobe.com/tiobe-index/. Accessed: 2025-02-20

    2025

  28. [37]

    Edward Suh

    Mohamed Ismail and G. Edward Suh. 2018. Quantitative Overhead Analysis for Python. In 2018 IEEE International Symposium on Workload Characterization (IISWC). 36–47. https://doi.org/10.1109/IISWC.2018.8573512

    2018

  29. [38]

    Jayavardhana Gubbi, Rajkumar Buyya, Slaven Marusic, and Marimuthu Palaniswami. 2013. Internet of Things (IoT): A vision, architectural elements, and future directions. Future Generation Computer Systems 29, 7 (2013), 1645–1660. https://doi.org/10.1016/j.future.2013.01.010 Incl...

    2013 doi

  30. [39]

    Bojan Karlaš, Babak Salimi, and Sebastian Schelter. 2024. Navigating Data Errors in Machine Learning Pipelines: Identify, Debug, and Learn. In Proceedings of SIGMOD’25. Association for Computing Machinery, New York, NY, USA

    2024

  31. [40]

    Mes, and Maria E

    Martijn Koot, Martijn R.K. Mes, and Maria E. Iacob. 2021. A systematic literature review of supply chain decision making supported by the Internet of Things and Big Data Analytics. Computers & Industrial Engineering 154 (April 2021), 107076. https://doi.org/10.1016/j.cie.2020.107076

    2021

  32. [41]

    Kaggle. [n. d.]. New York City Taxi Trip Duration Competition. https://www. kaggle.com/competitions/nyc-taxi-trip-duration. Accessed: 2025-05-27. V. Papastergios and A. Gounaris

    2025

  33. [42]

    Jin Li, David Maier, Kristin Tufte, Vassilis Papadimos, and Peter A. Tucker. 2005. Semantics and evaluation techniques for window aggregates in data streams. In Proceedings of the 2005 ACM SIGMOD International Conference on Management of Data (Baltimore, Maryland) (SIGMOD ’05)...

    2005

  34. [43]

    Python Pandas Library. [n. d.]. Python Pandas Library Official Website and Documentation. https://pandas.pydata.org/. Accessed: 2025-02-20

    2025

  35. [44]

    Franklin, and Ken Gold- berg

    Sanjay Krishnan, Jiannan Wang, Eugene Wu, Michael J. Franklin, and Ken Gold- berg. 2016. ActiveClean: interactive data cleaning for statistical modeling. Proc. VLDB Endow. 9, 12 (Aug. 2016), 948–959. https://doi.org/10.14778/2994509. 2994514

    2016 doi

  36. [45]

    Dragut, Arjun Mukherjee, and Weiyi Meng

    Qingyuan Liu, Eduard C. Dragut, Arjun Mukherjee, and Weiyi Meng. 2015. FLORIN: a system to support (near) real-time applications on user generated content on daily news. Proc. VLDB Endow. 8, 12 (Aug. 2015), 1944–1947. https: //doi.org/10.14778/2824032.2824107

    2015

  37. [46]

    David Loshin. 2011. Business Impacts of Poor Data Quality . Elsevier, 1–16. https://doi.org/10.1016/b978-0-12-373717-5.00001-4

    2011 doi

  38. [47]

    David Linthicum. 2024. Why companies are leaving the cloud. https://www. infoworld.com/article/2336102/why-companies-are-leaving-the-cloud.html. Ac- cessed: 18/01/2025

    2024

  39. [48]

    Sedir Mohammed, Lou Therese Brandner, Felicia Burtscher, Sebastian Hal- lensleben, Hazar Harmouch, Andreas Hauschke, Jessica Heesen, Stefanie Hilde- brandt, Simon David Hirsbrunner, Julia Keselj, Philipp Mahlow, Marie Massow, Felix Naumann, Frauke Rostalski, Anna Wilken, and A...

    2024 doi

  40. [49]

    Sedir Mohammed, Lisa Ehrlinger, Hazar Harmouch, Felix Naumann, and Divesh Srivastava. 2024. The Five Facets of Data Quality Assessment. arXiv:2403.00526 [cs.DB] https://arxiv.org/abs/2403.00526

    2024 arXiv

  41. [50]

    David Maier, Jin Li, Peter Tucker, Kristin Tufte, and Vassilis Papadimos. 2004. Semantics of Data Streams and Operators . Springer Berlin Heidelberg, 37–52. https://doi.org/10.1007/978-3-540-30570-5_3

    2004 doi

  42. [51]

    Vasileios Papastergios and Anastasios Gounaris. 2024. A survey of open-source data quality tools: shedding light on the materialization of data quality dimen- sions in practice. CoRR abs/2407.18649 (2024). https://doi.org/10.48550/ARXIV. 2407.18649

    2024 doi arXiv

  43. [52]

    Yeonsu Park, Byungchul Tak, and Wook-Shin Han. 2023. QaaD (Query-as-a- Data): Scalable Execution of Massive Number of Small Queries in Spark. Proc. ACM Manag. Data 1, 2, Article 134 (June 2023), 26 pages. https://doi.org/10.1145/ 3589279

    2023

  44. [53]

    PYPL PopularitY of Programming Language. [n. d.]. PYPL PopularitY of Pro- gramming Language. https://pypl.github.io/PYPL.html. Accessed: 2025-02-20

    2025

  45. [54]

    Rana, Sheshadri Chatterjee, Yogesh K

    Nripendra P. Rana, Sheshadri Chatterjee, Yogesh K. Dwivedi, and Shahriar Akter

  46. [55]

    Thomas C. Redman. 1998. The impact of poor data quality on the typical enter- prise. Commun. ACM 41, 2 (Feb. 1998), 79–82. https://doi.org/10.1145/269012. 269025

    1998 doi

  47. [56]

    Yuanhui Qiu, Chenguang Fang, Shaoxu Song, Xiangdong Huang, Chen Wang, and Jianmin Wang. 2023. TsQuality: Measuring Time Series Data Quality in Apache IoTDB. Proceedings of the VLDB Endowment 16, 12 (2023), 3982–3985

    2023

  48. [57]

    Apache Griffin GitHub Repository. 2025. Apache Griffin Measures for Stream- ing Data. https://github.com/apache/griffin/blob/master/griffin-doc/measure/ measure-streaming-sample.md

    2025

  49. [58]

    Bytewax GitHub Repository. 2024. Bytewax GitHub Repository. https://github. com/bytewax/bytewax

    2024

  50. [59]

    Deequ GitHub Repository. 2025. Deequ GitHub Repository. https://github.com/ awslabs/deequ

    2025

  51. [60]

    Apache Griffin GitHub Repository. 2025. Apache Griffin GitHub Repository. https://github.com/apache/griffin

    2025

  52. [61]

    Faust (Mirror) GitHub Repository. 2025. Faust GitHub Repository. https://github. com/faust-streaming/faust

    2025

  53. [62]

    Great Expectations GitHub Repository. 2025. Great Expectations GitHub Reposi- tory. https://github.com/great-expectations/great_expectations

    2025

  54. [63]

    MobyDQ GitHub Repository. 2025. MobyDQ GitHub Repository. https://github. com/ubisoft/mobydq

    2025

  55. [64]

    Evidently GitHub Repository. 2025. Evidently GitHub Repository. https://github. com/evidentlyai/evidently

    2025

  56. [65]

    Quix Streams GitHub Repository. 2025. Quix Streams GitHub Repository. https: //github.com/quixio/quix-streams

    2025

  57. [66]

    Soda Core GitHub Repository. 2025. Soda Core GitHub Repository. https: //github.com/sodadata/soda-core

    2025

  58. [67]

    Vincent Restat, Meike Klettke, and Uta Störl. 2023. "FAIR" is not enough–A Metrics Framework to ensure Data Quality through Data Preparation. In BTW

    2023

  59. [68]

    Pathway GitHub Repository. 2025. Pathway GitHub Repository. https://github. com/pathwaycom/pathway

    2025

  60. [69]

    Sebastian Schelter, Dustin Lange, Philipp Schmidt, Meltem Celikel, Felix Biess- mann, and Andreas Grafberger. 2018. Automating Large-Scale Data Quality Verification. Proceedings of the VLDB Endowment 11, 12 (2018), 1781–1794

    2018

  61. [70]

    Christoph Schinninger, Fabian Panse, Constantin Kühne, and Lisa Ehrlinger

  62. [71]

    Laura Sebastian-Coleman. 2013. Measuring Data Quality for Ongoing Improve- ment: A Data Quality Assessment Framework . Elsevier, Waltham, MA, USA

    2013

  63. [72]

    https://doi.org/10.18420/BTW2023-61

    doi

  64. [73]

    Sebastian Schelter, Stefan Grafberger, Philipp Schmidt, Tammo Rukat, Mario Kiessling, Andrey Taptunov, Felix Biessmann, and Dustin Lange. 2019. Differen- tial Data Quality Verification on Partitioned Data. In2019 IEEE 35th International Conference on Data Engineering (ICDE). 1...

    2019

  65. [74]

    Phanwadee Sinthong, Dhaval Patel, Nianjun Zhou, Shrey Shrivastava, Arun Iyen- gar, and Anuradha Bhamidipaty. 2022. DQDF: Data-Quality-Aware Dataframes. Proceedings of the VLDB Endowment 15, 4 (2022), 949–957

    2022

  66. [75]

    Kalashnikov, and Felix Naumann

    Divesh Srivastava, Tobias Bleifuß, Leon Bornemann, Dmitri V. Kalashnikov, and Felix Naumann. 2022. Exploring and Analyzing Change: The Janus Project. In Proceedings of the 31st ACM International Conference on Information & Knowl- edge Management (Atlanta, GA, USA) (CIKM ’22). ...

    2022

  67. [76]

    Strong, Yang W

    Diane M. Strong, Yang W. Lee, and Richard Y. Wang. 1997. Data quality in context. Commun. ACM 40, 5 (May 1997), 103–110. https://doi.org/10.1145/253769.253804

    1997

  68. [77]

    James M. Tien. 2017. Internet of Things, Real-Time Decision Making, and Ar- tificial Intelligence. Annals of Data Science 4, 2 (May 2017), 149–178. https: //doi.org/10.1007/s40745-017-0112-5

    2017 doi

  69. [78]

    Flavia Serra, Verónika Peralta, Adriana Marotta, and Patrick Marcel. 2024. Use of Context in Data Quality Management: A Systematic Literature Review. J. Data and Information Quality 16, 3, Article 19 (Oct. 2024), 41 pages. https: //doi.org/10.1145/3672082

    2024 doi

  70. [79]

    Shrey Shrivastava, Dhaval Patel, Anuradha Bhamidipaty, Wesley M Gifford, Stuart A Siegel, Venkata Sitaramagiridharganesh Ganapavarapu, and Jayant R Kalagnanam. 2019. DQA: Scalable, Automated and Interactive Data Quality Advisor. In IEEE International Conference on Big Data (Bi...

    2019

  71. [80]

    Jiannan Wang, Xin Luna Dong, and Michael Stonebraker. 2018. The Data Civilizer System. In Conference on Innovative Data Systems Research (CIDR)

    2018

  72. [81]

    Jianwei Wang, Ying Zhang, Kai Wang, Xuemin Lin, and Wenjie Zhang. 2024. Missing Data Imputation with Uncertainty-Driven Network. Proc. ACM Manag. Data 2, 3, Article 117 (May 2024), 25 pages. https://doi.org/10.1145/3654920

    2024 doi

  73. [82]

    Richard Y. Wang. 1998. A Product Perspective on Total Data Quality Management. Commun. ACM 41, 2 (1998), 58–65

    1998

  74. [83]

    Wang and Diane M

    Richard Y. Wang and Diane M. Strong. 1996. Beyond Accuracy: What Data Quality Means to Data Consumers. Journal of Management Information Systems 12 (1996), 5–33. Issue 4

    1996

  75. [84]

    Ankit Toshniwal, Siddarth Taneja, Amit Shukla, Karthik Ramasamy, Jignesh M Patel, Sanjeev Kulkarni, Jason Jackson, Krishna Gade, Maosong Fu, Jake Donham, et al. 2014. Storm@ twitter. In Proceedings of the 2014 ACM SIGMOD international conference on Management of data . 147–156

    2014

  76. [85]

    Rajeshwari U and B Sathish Babu. 2016. Real-time credit card fraud detection using Streaming Analytics. In 2016 2nd International Conference on Applied and Theoretical Computing and Communication Technology (iCATccT). 439–444. https: //doi.org/10.1109/ICATCCT.2016.7912039

    2016

  77. [86]

    Matei Zaharia, Tathagata Das, Haoyuan Li, Timothy Hunter, Scott Shenker, and Ion Stoica. 2013. Discretized streams: fault-tolerant streaming computation at scale. In Proceedings of the Twenty-Fourth ACM Symposium on Operating Systems Principles (Farminton, Pennsylvania) (SOSP ...

    2013

  78. [90]

    Apache Griffin Official Website. 2025. Streaming Use Cases. https://griffin.apache. org/docs/usecases.html

    2025

  79. [91]

    Matei Zaharia, Mosharaf Chowdhury, Tathagata Das, Ankur Dave, Justin Ma, Murphy McCauley, Michael Franklin, Scott Shenker, and Ion Stoica. 2012. Re- silient distributed datasets: A fault-tolerant abstraction for in-memory cluster computing. Proceedings of the 9th USENIX Confer...

    2012

  80. [2019]

    In Proceedings of SysML

    Data Validation for Machine Learning. In Proceedings of SysML . https: //mlsys.org/Conferences/2019/doc/2019/167.pdf

    2019

  81. [2021]

    2021), 364–387

    Understanding dark side of artificial intelligence (AI) integrated business analytics: assessing firm’s operational inefficiency and competitiveness.European Journal of Information Systems 31, 3 (Aug. 2021), 364–387. https://doi.org/10. 1080/0960085x.2021.1955628

    2021

  82. [2023]

    ACM Comput

    Landscape of High-Performance Python to Develop Data Science and Machine Learning Applications. ACM Comput. Surv. 56, 3, Article 65 (Oct. 2023), 30 pages. https://doi.org/10.1145/3617588

    2023 doi

  83. [2025]

    https://doi.org/10.48786/ EDBT.2025.64

    Icewafl: A Configurable Data Stream Polluter. https://doi.org/10.48786/ EDBT.2025.64

    2025

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.