Pith. sign in

REVIEW 4 major objections 5 minor 127 references

The Impact of Modern AI in Metadata Management

T0 review · 4 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read A modular AI framework can automate metadata management at scale.

desk verdict A useful survey of metadata tools with a framework that is honestly labeled conceptual in the body, but the abstract overstates it as a solution. read the letter →

arxiv 2501.16605 v2 pith:KEDTWVWP submitted 2025-01-28 cs.DB cs.AI

Wenli Yang , Rui Fu , Muhammad Bilal Amin , Byeong Kang This is my paper · ORCID
classification cs.DBcs.AI
keywords metadatamanagementAI-drivenautomatedgenerationdatagovernancelargelanguagemodelsknowledgegraphsnext-generationdatasetscatalog
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

This paper argues that metadata management is at a turning point: traditional approaches such as manual cataloguing and rule-based systems cannot scale to modern datasets, and existing AI-powered tools, while better, still struggle with data dependency, limited interpretability, ambiguity, lifecycle coverage, and inconsistency. The paper surveys open-source and commercial metadata systems, maps AI techniques onto the main metadata modules, and proposes a conceptual AI-assisted framework that combines automated metadata generation, quality assurance and governance, and advanced analytics and accessibility. The claim is that such a modular framework, built on machine learning, deep learning, natural language processing, knowledge graphs, and large language models, would let organizations automate metadata creation, enforce governance, and keep datasets usable as they grow in volume and complexity. A sympathetic reader would care because the paper offers a reference model for turning AI capability into practical metadata management, even though the framework itself is conceptual and untested.

What carries the argument

The load-bearing object is the proposed AI-assisted metadata management framework, a conceptual reference architecture whose main modules are automated metadata generation, quality assurance and governance, and advanced analytics and accessibility. It is load-bearing because the paper's strongest claim is carried by this architecture rather than by an implemented system. The framework's work is to show how machine learning, deep learning, natural language processing, knowledge graphs, and large language models could be assembled as modular services, connected by APIs, to automate metadata creation, validate and govern metadata, and deliver insight through analytics and visualization.

What would settle it

Run a controlled evaluation where AI-generated metadata is produced with the framework's design and compared against expert-created metadata for the same heterogeneous documents, measuring field-level precision and recall plus the time users need to find a requested dataset; if the AI-generated metadata is not at least as accurate and takes as long to correct as manually created metadata, the claimed automation benefit collapses.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is a structured picture of how AI is, and could be, used in metadata management, along with a proposed reference architecture that maps six metadata functions to specific AI techniques: extraction and generation, search and discovery, quality management, storage and indexing, lineage and governance, and collaboration and socialization. The survey shows that no single tool or technique covers all these functions well: language-based AI is strong at extraction and classification, knowledge-centred methods such as knowledge graphs support reasoning but are hard to scale, and generative AI and large language models offer broad coverage but weak explainability. The paper claims that integrating these techniques in one framework, with API-driven interoperability, automated validation, and governance, would automate metadata generation and improve the accessibility and usability of next-generation datasets.

Load-bearing premise

The framework assumes that current AI techniques, especially large language models and knowledge graphs, can actually be integrated into the proposed modules and will deliver the promised automation and governance benefits in real deployments.

Editorial extensions

If this is right

  • Organizations can use the proposed architecture as a blueprint: metadata creation becomes automated, governance becomes embedded policy checks, and users get analytics and visualization over metadata.
  • No single AI technique covers all metadata functions; deployment should pair language models for extraction with knowledge-centred methods for lineage and reasoning, plus rule-based validation for quality.
  • Next-generation datasets, which are large, heterogeneous, and fast-moving, are the clearest beneficiaries because the framework is specifically pitched at their scale and complexity.
  • The paper's functional-capability comparison gives a common yardstick to assess both open-source and commercial metadata tools module by module.
  • Future work implied by the paper includes continual learning for streaming metadata, energy-efficient edge deployment, and decentralized audit trails for compliance.

Reading between the lines

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

  • Editorial inference: the framework's concrete value will hinge on measuring whether AI-generated metadata passes expert review without correction; the paper does not report such measurements.
  • Editorial inference: the capability matrix can be read as a prescriptive decision tool, choosing knowledge graphs for lineage and reasoning, generative AI for extraction and enrichment, and rule-based checks for validation, so no single model carries all modules.
  • Editorial inference: a natural next experiment is to use the framework's validation module as a benchmark harness, comparing AI-generated metadata against hand-curated metadata on precision, recall, and user retrieval time across heterogeneous datasets.
  • Editorial inference: if the framework works as intended, metadata management shifts from archiving to continuous curation, connecting with designs that treat metadata as a living, frequently updated product rather than a static record.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

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. This paper surveys metadata management from traditional tools (Amundsen, CKAN, Meta-Grid, and an 'Atlas/Atlan' entry) through AI-driven open-source and commercial platforms (DataHub, OpenMetadata, Alation, Collibra, Informatica, and others), and provides two comparative analyses: one of AI techniques used across metadata management modules (Table 5) and one of functional capabilities such as extraction, classification, reasoning, and explainability (Table 6). It then identifies gaps in traditional and AI-powered approaches, especially data dependency, interpretability, ambiguity, lifecycle coverage, and validation, and proposes a conceptual AI-assisted metadata management framework (Section 4.2, Figure 3) aimed at automated metadata generation, governance, and accessibility. The final sections discuss future directions in scalable infrastructure, advanced AI, governance, interoperability, and collaborative validation.

Significance. If taken as a survey, the paper is useful: it consolidates a broad range of tools and techniques, and Tables 5 and 6 provide a structured comparison that practitioners and researchers could use to position new work. The paper is also honest in Section 4.2.1 and Section 4.2.3 in stating that the proposed framework is conceptual and that evaluation is future work. However, the abstract and conclusion go beyond this by presenting the framework as a 'promising solution' and 'designed to address these challenges,' a claim that is not supported by any implementation, prototype, pilot, or empirical evaluation. The framework's quality-assurance capabilities are asserted rather than designed, and the paper's own catalog of AI limitations is not mitigated by any described mechanism. The contribution is therefore a descriptive survey plus a high-level reference sketch, not a validated solution; the framing must be aligned with that status.

major comments (4)
  1. [Section 6 and Abstract] The conclusion states that 'The proposed AI-assisted metadata management framework offers a promising solution to these challenges,' and the abstract describes the framework as 'designed to address these challenges.' This is not supported by the manuscript. Section 4.2.1 explicitly says the framework is 'conceptual' and 'rather than presenting a fully operational system,' and Section 4.2.3 says future research 'will focus on evaluating its performance through empirical case studies.' Moreover, Section 4.1.2 lists data dependency, limited interpretability, ambiguity, lifecycle gaps, and inconsistency as limitations of existing AI solutions, and Table 5 notes that LLMs 'may lack transparency' and GenAI 'may introduce errors or inconsistencies if not properly monitored'; the proposed framework does not explain how it mitigates these limitations. The central claim should be recalibrated to present the contribution as a conceptual reference architecture with open validation questions, or the authors should add module-level mechanisms that concretely address the listed limitations.
  2. [Section 4.2.2, Figure 3] The key capabilities 'Metadata Validation & Verification' and 'Consistency & Compliance' are asserted, but no design is given for how they would work. The high-level architecture in Figure 3 does not identify components or interfaces for detecting or correcting inconsistent LLM or GenAI output, despite Table 5's warning that generative models 'may introduce errors or inconsistencies if not properly monitored.' Without such mechanisms, the framework's claim to enhance governance and ensure trustworthy metadata is not substantiated. The authors should either specify a validation and verification design (e.g., rule-based checkpoints, human-in-the-loop review, cross-source reconciliation, or confidence scoring) or explicitly mark these as open problems to be addressed in future work.
  3. [Section 3.1.2] The taxonomy is inconsistent and factually problematic around the 'Atlas' entry. Section 3.1.2 is titled 'Commercial Metadata Tools' and says 'Atlas is a commercial solution,' but the tool described in the text is Atlan, a commercial data catalog, while Apache Atlas is an open-source Apache project. This mischaracterization matters because the paper's contribution includes a comparative analysis of traditional versus AI-driven tools; an incorrect classification threatens the reliability of that comparison. The authors should correct the naming, distinguish Apache Atlas from Atlan, and place each tool in the appropriate section.
  4. [Section 3.2.1] DataGalaxy is listed as an 'AI-Driven Open-Source Platform,' but DataGalaxy is generally marketed as a proprietary, closed-source SaaS data catalog. If the authors have evidence that it is open-source, that evidence should be cited; otherwise, it should be moved to the commercial tools section or removed from the open-source list. The same verification should be applied to the other tools in Table 3 to avoid repeating a classification error in a paper whose central survey value depends on accurate taxonomy.
minor comments (5)
  1. [Section 2.2] The opening sentence, 'Metadata originates from multiple sources and file types for storage, including drug management, patient status management, geographic information...' is awkward; consider rephrasing to 'Metadata is drawn from diverse source domains, including...' for clarity.
  2. [Table 6] The legend describes strong, partial, and no support as symbols, but the symbols are not visible in the text version; ensure they render correctly in the published PDF and are also described in words for accessibility.
  3. [References] References [56] and [58] appear to be duplicates of the same paper ('From text to insight: large language models for materials science data extraction'); consolidate them into a single citation.
  4. [Section 4.2.1] The phrase 'It offers a scalable, adaptable, and secure architecture' is stated as fact, but scalability and security are not demonstrated anywhere in the manuscript; suggest changing to 'intended to offer' or 'designed to support' to match the conceptual status.
  5. [Section 5.2.2 and 5.3.2] Several future-direction bullets use 'will integrate' and 'will adopt' for features that are not yet part of the framework; consider using conditional or exploratory language (e.g., 'we plan to investigate') to avoid implying these components already exist.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the paper is a survey plus explicitly conceptual framework, with no derived predictions, fitted parameters, or load-bearing self-citations.

full rationale

The paper contains no equations, no fitted parameters, and no derived prediction whose value could reduce to an input by construction. Its contribution is a comparative survey of traditional and AI-driven metadata tools, followed by a conceptual reference architecture (Figure 3) whose capabilities are described qualitatively. The framework's claims are explicitly qualified as a blueprint rather than an implemented system: Section 4.2.1 states it is 'a conceptual framework designed to serve as both a reference model and a blueprint for prototyping AI-powered metadata systems for future implementation and evaluation,' and Section 4.2.3 states 'While the proposed framework is conceptual... Future research will focus on evaluating its performance through empirical case studies.' There are no self-citations that carry the argument; the cited references are external works on metadata standards, AI techniques, and tools. The observation that the framework is presented as addressing gaps the authors themselves define is a completeness or validation concern, not a circular derivation, because no claimed result is logically equivalent to its own input. Accordingly, no significant circularity is present.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

The paper is a survey and conceptual proposal, so it introduces no mathematical free parameters and no new entities. It does, however, rest on unverified domain assumptions about tool taxonomies and the practical effectiveness of AI techniques, plus the implicit feasibility of the proposed framework.

assumptions (3)
  • domain assumption The decomposition of metadata management into the six modules in Section 2.3 and the classification of tools into traditional vs AI-driven are a faithful representation of the field.
    The survey's comparisons and gap analysis rely on this taxonomy, which is presented without a systematic methodology for tool selection or module definition.
  • ad hoc to paper The AI techniques listed in Table 5 can be applied to the corresponding metadata modules and will provide the stated advantages in practice.
    The proposed framework assumes these techniques deliver the described benefits, while Section 4.1.2 acknowledges unresolved limitations such as interpretability and data dependency.
  • ad hoc to paper The proposed framework's architecture can be implemented with existing technology and integrated with LLMs, knowledge graphs, and other AI services.
    No prototype or feasibility study is supplied; this is the central unverified premise of the proposal.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The Impact of Modern AI in Metadata Management." pith.science (2026). https://pith.science/paper/KEDTWVWP

@misc{pith2026250116605,
  author       = {Pith},
  title        = {Pith review of: The Impact of Modern AI in Metadata Management},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KEDTWVWP}},
  note         = {Machine review of arXiv:2501.16605}
}
read the original abstract

Metadata management plays a critical role in data governance, resource discovery, and decision-making in the data-driven era. While traditional metadata approaches have primarily focused on organization, classification, and resource reuse, the integration of modern artificial intelligence (AI) technologies has significantly transformed these processes. This paper investigates both traditional and AI-driven metadata approaches by examining open-source solutions, commercial tools, and research initiatives. A comparative analysis of traditional and AI-driven metadata management methods is provided, highlighting existing challenges and their impact on next-generation datasets. The paper also presents an innovative AI-assisted metadata management framework designed to address these challenges. This framework leverages more advanced modern AI technologies to automate metadata generation, enhance governance, and improve the accessibility and usability of modern datasets. Finally, the paper outlines future directions for research and development, proposing opportunities to further advance metadata management in the context of AI-driven innovation and complex datasets.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

127 extracted references · 80 canonical work pages

  1. [1]

    Understanding the nature of metadata: system - atic review

    Ulrich H, et al. Understanding the nature of metadata: system - atic review. J Med Internet Res. 2022;24(1):e25440

    2022

  2. [2]

    The role of metadata in reproducible compu- tational research

    Leipzig J, et al. The role of metadata in reproducible compu- tational research. Patterns. 2021;2(9):100322

    2021

  3. [3]

    Improving the documentation and findability of data services and repositories: a review of (meta) data manage- ment approaches

    Řezník T, et al. Improving the documentation and findability of data services and repositories: a review of (meta) data manage- ment approaches. Comput Geosci. 2022;169:105194

    2022

  4. [4]

    Metadata as data intelligence

    Greenberg J, et al. Metadata as data intelligence. Data Intell. 2023;5(1):1–5

    2023

  5. [5]

    Big data analytics capability and firm performance: meta-analysis

    Ansari K, Ghasemaghaei M. Big data analytics capability and firm performance: meta-analysis. J Comput Inform Syst. 2023;63(6):1477–94. Human-Centric Intelligent Systems

    2023

  6. [6]

    health care management

    Chowdhury RH. Big data analytics in the field of multifaceted analyses: a study on “health care management.” World J Adv Res Rev. 2024;22(3):2165–72

    2024

  7. [7]

    Trends and future perspective challenges in big data

    Naeem M, et al. Trends and future perspective challenges in big data. In: Proceeding of the sixth Euro-China conference on intel- ligent data analysis and applications. 2022. p. 309–25

    2022

  8. [8]

    Data mesh: a systematic gray literature review

    Goedegebuure A, et al. Data mesh: a systematic gray literature review. ACM Comput Surv. 2024;57(1):1–36

    2024

Show all 127 references
  1. [9]

    Re-thinking data strategy and integration for artificial intelligence: concepts, opportunities, and challenges

    Aldoseri A, Al-Khalifa KN, Hamouda AM. Re-thinking data strategy and integration for artificial intelligence: concepts, opportunities, and challenges. Appl Sci. 2023;13(12):7082

    2023

  2. [10]

    A review of the state of the art of data quality in healthcare

    Liu C, et al. A review of the state of the art of data quality in healthcare. J Glob Inform Manag. 2023;31(1):1–18

    2023

  3. [11]

    Data catalogs in the enterprise: applications and integration

    Jahnke N, Otto B. Data catalogs in the enterprise: applications and integration. Datenbank-Spektrum. 2023;23(2):89–96

    2023

  4. [12]

    Application of artificial intelligence (AI) in libraries and its impact on library operations review

    Subaveerapandiyan A. Application of artificial intelligence (AI) in libraries and its impact on library operations review. 2023. 10.6084/m9.figshare.22573345.v1

    2023 doi

  5. [13]

    Towards augmenting metadata management by machine learning

    Kern CJ, Schäffer T, Stelzer D. Towards augmenting metadata management by machine learning. In: INFORMATIK 2021

    2021

  6. [14]

    The role of AI in transforming metadata man- agement: insights on challenges, opportunities, and emerging trends

    Oyighan D, et al. The role of AI in transforming metadata man- agement: insights on challenges, opportunities, and emerging trends. Asian J Inform Sci Technol. 2024;14(2):20–6

    2024

  7. [15]

    Big data acquisition

    Lyko K, Nitzschke M, Ngonga Ngomo A-C. Big data acquisition. New horizons for a data-driven economy: a roadmap for usage and exploitation of big data in Europe. 2016: p. 39–61

    2016

  8. [16]

    Towards automated data cleaning workflows

    Mahdavi M, et al. Towards automated data cleaning workflows. Mach Learn. 2019;15:16

    2019

  9. [17]

    Metadata verification: a workflow for computa- tional archival science

    Pepper J, et al. Metadata verification: a workflow for computa- tional archival science. In: 2022 IEEE international conference on Big Data (Big Data). 2022. p. 2565–71

    2022

  10. [18]

    Metadata standard for continuous pres- ervation, discovery, and reuse of research data in repositories by higher education institutions: a systematic review

    Mosha NF, Ngulube P. Metadata standard for continuous pres- ervation, discovery, and reuse of research data in repositories by higher education institutions: a systematic review. Information. 2023;14(8):427

    2023

  11. [19]

    Metadata for digital libraries: state of the art and future directions

    Gartner R, L’Hours H, Young G. Metadata for digital libraries: state of the art and future directions. Bristol, UK: JISC; 2008

    2008

  12. [20]

    Metadata for digital collections

    Miller SJ. Metadata for digital collections. American Library Association; 2022

    2022

  13. [21]

    Improving social book search using structure semantics, bibliographic descriptions and social meta- data

    Ullah I, Khusro S, Ahmad I. Improving social book search using structure semantics, bibliographic descriptions and social meta- data. Multimedia Tools Appl. 2021;80(4):5131–72

    2021

  14. [22]

    Enhancing untargeted metabolomics using metadata-based source annotation

    Gauglitz JM, et al. Enhancing untargeted metabolomics using metadata-based source annotation. Nat Biotechnol. 2022;40(12):1774–9

    2022

  15. [23]

    Content management systems performance and compliance assessment based on a data-driven search engine optimization methodology

    Drivas I, et al. Content management systems performance and compliance assessment based on a data-driven search engine optimization methodology. Information. 2021;12(7):259

    2021

  16. [24]

    A strategy for archives metadata representation on CIDOC-CRM and knowledge dis- covery

    Melo D, Rodrigues IP, Varagnolo D. A strategy for archives metadata representation on CIDOC-CRM and knowledge dis- covery. Semantic Web. 2023;14(3):553–84

    2023

  17. [25]

    Metadata standards in web archiv- ing technological resources for ensuring the digital preservation of archived websites

    Formenton D, Gracioso LDS. Metadata standards in web archiv- ing technological resources for ensuring the digital preservation of archived websites. RDBCI Revista Digital de Biblioteconomia e Ciência da Informação. 2023;20:e022001

    2023

  18. [26]

    Computational metadata generation meth- ods for biological specimen image collections

    Karnani K, et al. Computational metadata generation meth- ods for biological specimen image collections. Int J Digit Libr. 2024;25(2):157–74

    2024

  19. [27]

    Githru: visual analytics for understanding software development history through git metadata analysis

    Kim Y, et al. Githru: visual analytics for understanding software development history through git metadata analysis. IEEE Trans Visual Comput Graphics. 2020;27(2):656–66

    2020

  20. [28]

    Metadata integration for spam reviews detection on Vietnamese e-commerce websites

    Van Dinh C, Luu ST. Metadata integration for spam reviews detection on Vietnamese e-commerce websites. Int J Asian Lang Process. 2024;34:245002. https:// doi. org/ 10. 1142/ S2717 55452 45000 24

    2024

  21. [29]

    Metadata concepts for advancing the use of digital health technologies in clinical research

    Badawy R, et al. Metadata concepts for advancing the use of digital health technologies in clinical research. Digital Biomark- ers. 2020;3(3):116–32

    2020

  22. [30]

    A metadata-assisted cascading ensemble clas- sification framework for automatic annotation of open IoT data

    Montori F, et al. A metadata-assisted cascading ensemble clas- sification framework for automatic annotation of open IoT data. IEEE Internet Things J. 2023;10(15):13401–13

    2023

  23. [31]

    Metadata management in data lake environments: a survey

    Boukraa D, Bala M, Rizzi S. Metadata management in data lake environments: a survey. J Libr Metadata. 2024;24(4):215–74

    2024

  24. [32]

    Metadata based classification techniques for knowledge discovery from facebook multimedia database

    Bhat P, Malaganve P. Metadata based classification techniques for knowledge discovery from facebook multimedia database. Int J Intell Syst Appl. 2021;13(4):38

    2021

  25. [33]

    Metadata quality in the era of big data and unstructured content

    Elouataoui W, El Alaoui I, Gahi Y. Metadata quality in the era of big data and unstructured content. In: Advances in information, communication and cybersecurity: proceedings of ICI2C’21

  26. [34]

    Efficient metadata indexing for hpc storage sys- tems

    Paul AK, et al. Efficient metadata indexing for hpc storage sys- tems. In: 20th IEEE/ACM international symposium on cluster, cloud and internet computing (CCGRID). 2020. p. 162–71

    2020

  27. [35]

    Literature review on metadata governance

    Kaur A, et al. Literature review on metadata governance. Open Int J Inform. 2023;11(1):114–20

    2023

  28. [36]

    The collaborative metadata repository (CoMetaR) web app: quantitative and qualitative usability evaluation

    Stöhr MR, Günther A, Majeed RW. The collaborative metadata repository (CoMetaR) web app: quantitative and qualitative usability evaluation. JMIR Med Inform. 2021;9(11):e30308

    2021

  29. [37]

    Hands off the metadata!: comparing the use of explicit and background metadata in crowdsourced dialectol- ogy

    Blaxter T, Britain D. Hands off the metadata!: comparing the use of explicit and background metadata in crowdsourced dialectol- ogy. Linguistics Vanguard. 2021;7(s1):20190029

    2021

  30. [38]

    Information experiences of organisational newcomers: using public social media for organisational socialisation

    Huang V. Information experiences of organisational newcomers: using public social media for organisational socialisation. Behav Inform Technol. 2023;42(9):1279–93

    2023

  31. [39]

    Dublin Core’s DCMIType ‘PhysicalObject’ and its use across the open language archives community

    Paterson III H. Dublin Core’s DCMIType ‘PhysicalObject’ and its use across the open language archives community. In: Proceedings of the 17th annual society of American archivists research forum. 2023

    2023

  32. [40]

    Hilbring D. et al. OData-usage of a REST based API standard in web based environmental information systems. In: EnviroInfo

  33. [41]

    Accessible search and the role of meta- data

    Beyene WM, Godwin T. Accessible search and the role of meta- data. Library Hi Tech. 2018;36(1):2–17

    2018

  34. [42]

    Building a multitenant data hub system using elastic stack and kafka for uniform data representation

    Kuduz N, Salapura S. Building a multitenant data hub system using elastic stack and kafka for uniform data representation. In: 19th international symposium INFOTEH-JAHORINA (INFOTEH). 2020. p. 1–6

    2020

  35. [43]

    DataHub and apache atlas: a comparative analysis of data catalog tools

    Rodrigues D, et al. DataHub and apache atlas: a comparative analysis of data catalog tools. In: CAPSI 2022 Proceedings

    2022

  36. [44]

    Quality assessment of open datasets metadata

    Šlibar B. Quality assessment of open datasets metadata. Univer- sity of Zagreb; 2024

    2024

  37. [45]

    Metadata extraction using semantic and natural language processing techniques

    Knapen R, et al. Metadata extraction using semantic and natural language processing techniques. In: iEMSs conference. 2014. p. 48

    2014

  38. [46]

    Rule based metadata extraction frame- work from academic articles

    Azimjonov J, Alikhanov J. Rule based metadata extraction frame- work from academic articles. 2018. https:// doi. org/ 10. 48550/ arXiv. 1807. 09009

    2018

  39. [47]

    Cleaning by clustering: methodology for address- ing data quality issues in biomedical metadata

    Hu W, et al. Cleaning by clustering: methodology for address- ing data quality issues in biomedical metadata. BMC Bioinform. 2017;18:1–12

    2017

  40. [48]

    Automatic extraction and cluster analysis of natu- ral disaster metadata based on the unified metadata framework

    Wang Z, et al. Automatic extraction and cluster analysis of natu- ral disaster metadata based on the unified metadata framework. ISPRS Int J Geo Inf. 2024;13(6):201

    2024

  41. [49]

    Document classification based on meta- data and keywords extraction

    Rezqa EY, Baraka RS. Document classification based on meta- data and keywords extraction. In: Palestinian international conference on information and communication technology (PICICT). 2021. p. 18–24

    2021

  42. [50]

    FLAG-PDFe: Features oriented meta- data extraction framework for scientific publications

    Ahmed MW, Afzal MT. FLAG-PDFe: Features oriented meta- data extraction framework for scientific publications. IEEE Access. 2020;8:99458–69. Human-Centric Intelligent Systems

    2020

  43. [51]

    Suicidality detection on social media using meta- data and text feature extraction and machine learning

    Jung W, et al. Suicidality detection on social media using meta- data and text feature extraction and machine learning. Arch Sui- cide Res. 2023;27(1):13–28

    2023

  44. [52]

    Automatic metadata extraction incorpo- rating visual features from scanned electronic theses and dis - sertations

    Choudhury MH, et al. Automatic metadata extraction incorpo- rating visual features from scanned electronic theses and dis - sertations. In: ACM/IEEE joint conference on digital libraries (JCDL). 2021. p. 230–33

    2021

  45. [53]

    Deep neural networks-based classifi- cation methodologies of speech, audio and music, and its integra- tion for audio metadata tagging

    Park H, Chung Y, Kim J-H. Deep neural networks-based classifi- cation methodologies of speech, audio and music, and its integra- tion for audio metadata tagging. J Web Eng. 2023;22(1):1–26

    2023

  46. [54]

    Automatic document metadata extraction based on deep networks

    Liu R, et al. Automatic document metadata extraction based on deep networks. In: Natural language processing and Chinese computing: 6th CCF international conference. 2018. p. 305–17

    2018

  47. [55]

    CrossDomain recommendation based on MetaData using graph convolution networks

    Khan R, et al. CrossDomain recommendation based on MetaData using graph convolution networks. IEEE Access. 2023;11:90724–38

    2023

  48. [56]

    From text to insight: large language models for materials science data extraction

    Schilling-Wilhelmi M, et al. From text to insight: large language models for materials science data extraction. arXiv preprint arXiv: 2407. 16867, 2024

    2024

  49. [57]

    Impact of conversational and generative AI sys- tems on libraries: a use case large language model (LLM)

    Khan R, et al. Impact of conversational and generative AI sys- tems on libraries: a use case large language model (LLM). Sci Technol Libr. 2024;43(4):319–33

    2024

  50. [58]

    From text to insight: large language models for materials science data extraction

    Schilling-Wilhelmi M, et al. From text to insight: large language models for materials science data extraction. 2024. https:// doi. org/ 10. 48550/ arXiv. 2407. 16867

    2024

  51. [59]

    MatSciBERT: A materials domain language model for text mining and information extraction

    Gupta T, et al. MatSciBERT: A materials domain language model for text mining and information extraction. NPJ Comput Mater. 2022;8(1):102

    2022

  52. [60]

    Multi-task reinforcement learning with context-based representations

    Sodhani S, Zhang A, Pineau J. Multi-task reinforcement learning with context-based representations. In: International conference on machine learning. 2021. p. 9767–79

    2021

  53. [61]

    Extracting enhanced artificial intelligence model metadata from software repositories

    Tsay J, et al. Extracting enhanced artificial intelligence model metadata from software repositories. Empir Softw Eng. 2022;27(7):176

    2022

  54. [62]

    Design and data mining techniques for large-scale scholarly digital libraries and search engines

    Rohatgi S. Design and data mining techniques for large-scale scholarly digital libraries and search engines. The Pennsylvania State University; 2023

    2023

  55. [63]

    Embedding metadata using deep collaborative filtering to address the cold start problem for the rating prediction task

    Nahta R, et al. Embedding metadata using deep collaborative filtering to address the cold start problem for the rating prediction task. Multim Tools Appl. 2021;80:18553–81

    2021

  56. [64]

    Metadata-driven error detection

    Visengeriyeva L, Abedjan Z. Metadata-driven error detection. In: Proceedings of the 30th international conference on scientific and statistical database management. 2018. p. 1–12

    2018

  57. [65]

    Exploring dimensions of metadata quality assessment: a scoping review

    Kumar V, Chandrappa, Harinarayana N. Exploring dimensions of metadata quality assessment: a scoping review. J Librarianship Inform Sci. 2024. https:// doi. org/ 10. 1177/ 09610 00624 12390 80

    2024

  58. [66]

    A rule-based data quality assessment sys- tem for electronic health record data

    Wang Z, et al. A rule-based data quality assessment sys- tem for electronic health record data. Appl Clin Inform. 2020;11(04):622–34

    2020

  59. [67]

    Repairing raw metadata for metadata man- agement

    Khalid H, Zimányi E. Repairing raw metadata for metadata man- agement. Inf Syst. 2024;122:102344

    2024

  60. [68]

    Quality prediction of open educational resources a metadata-based approach

    Tavakoli M, et al. Quality prediction of open educational resources a metadata-based approach. In: IEEE 20th international conference on advanced learning technologies (ICALT). 2020. p. 29–31

    2020

  61. [69]

    A deep-learning based citation count prediction model with paper metadata semantic features

    Ma A, et al. A deep-learning based citation count prediction model with paper metadata semantic features. Scientometrics. 2021;126(8):6803–23

    2021

  62. [70]

    Open government data: usage trends and metadata quality

    Quarati A. Open government data: usage trends and metadata quality. J Inf Sci. 2023;49(4):887–910

    2023

  63. [71]

    A generic and customiz- able genetic algorithms-based conceptual model modularization framework

    Ali SJ, Michael Laranjo J, Bork D. A generic and customiz- able genetic algorithms-based conceptual model modularization framework. In: International conference on enterprise design, operations, and computing. 2023. p. 39–57

    2023

  64. [72]

    AI-Driven frameworks for enhancing data quality in big data ecosystems: Error_detection, correction, and metadata integration

    Elouataoui W. AI-Driven frameworks for enhancing data quality in big data ecosystems: Error_detection, correction, and metadata integration. 2024. https:// doi. org/ 10. 48550/ arXiv. 2405. 03870

    2024

  65. [73]

    The role of metadata in promoting explainabil- ity and interoperability of AI-based prediction models

    Ahmed AA, et al. The role of metadata in promoting explainabil- ity and interoperability of AI-based prediction models. J Except Multidiscip Res. 2024;1(1):33–45

    2024

  66. [74]

    Fuzzy metadata strategies for enhanced data integration

    Khalid H, Zimanyi E, Wrembel R. Fuzzy metadata strategies for enhanced data integration. In: Proceedings of the 7th interna- tional conference on data science, technology and applications

  67. [75]

    A framework for creating knowledge graphs of scientific software metadata

    Kelley A, Garijo D. A framework for creating knowledge graphs of scientific software metadata. Quant Sci Stud. 2021;2(4):1423–46

    2021

  68. [76]

    Knowledge graph quality management: a comprehensive survey

    Xue B, Zou L. Knowledge graph quality management: a comprehensive survey. IEEE Trans Knowl Data Eng. 2022;35(5):4969–88

    2022

  69. [77]

    MetaQA: enhancing human-centered data search using Generative Pre-trained Transformer (GPT) language model and artificial intelligence

    Li D, Zhang Z. MetaQA: enhancing human-centered data search using Generative Pre-trained Transformer (GPT) language model and artificial intelligence. PLoS ONE. 2023;18(11):e0293034

    2023

  70. [78]

    Metagraph: indexing and analysing nucleo- tide archives at petabase-scale

    Karasikov M, et al. Metagraph: indexing and analysing nucleo- tide archives at petabase-scale. BioRxiv. 2020. p. 2020. https:// doi. org/ 10. 1101/ 2020. 10. 01. 322164

    2020

  71. [79]

    Diesel: a dataset-based distributed storage and caching system for large-scale deep learning training

    Wang L, et al. Diesel: a dataset-based distributed storage and caching system for large-scale deep learning training. In: Pro- ceedings of the 49th international conference on parallel process- ing. 2020. p. 1–11

    2020

  72. [80]

    Machine learning and ontology-based novel semantic document indexing for information retrieval

    Sharma A, Kumar S. Machine learning and ontology-based novel semantic document indexing for information retrieval. Comput Ind Eng. 2023;176:108940

    2023

  73. [81]

    Metadata management and application

    Satija M, Bagchi M, Martínez-Ávila D. Metadata management and application. Libr Her. 2020;58(4):84–107

    2020

  74. [82]

    Building semantic metadata for historical archives through an ontology-driven user interface

    Goy A, et al. Building semantic metadata for historical archives through an ontology-driven user interface. J Comput Cult Herit. 2020;13(3):1–36

    2020

  75. [83]

    Ontology-supported AI model and dataset man- agement

    Novacek J, et al. Ontology-supported AI model and dataset man- agement. In: IEEE 22nd international conference on industrial informatics (INDIN). 2024. p. 1–6

    2024

  76. [84]

    Colt: concept lineage tool for data flow metadata capture and analysis

    Aggour KS, et al. Colt: concept lineage tool for data flow metadata capture and analysis. Proc VLDB Endow. 2017;10(12):1790–801

    2017

  77. [85]

    Optimizing data governance through AI-driven meta- data management: enhancing data discovery and utilization in organizations

    Li M-L. Optimizing data governance through AI-driven meta- data management: enhancing data discovery and utilization in organizations. Innovat Eng Sci J. 2022;2(1)

    2022

  78. [86]

    In: Data fabric and data mesh approaches with AI: a guide to AI-based data cataloging, governance, integration, orchestration, and consumption

    Hechler E, Weihrauch M, Wu Y, Intelligent cataloging and meta- data management. In: Data fabric and data mesh approaches with AI: a guide to AI-based data cataloging, governance, integration, orchestration, and consumption. Springer; 2023. p. 293–310

    2023

  79. [87]

    Implementing a block- chain-powered metadata catalog in data mesh architecture

    Dolhopolov A, Castelltort A, Laurent A. Implementing a block- chain-powered metadata catalog in data mesh architecture. In: International congress on blockchain and applications. 2023. p. 348–60

    2023

  80. [88]

    Integrating metadata into deep autoencoder for handling prediction task of collaborative recommender system

    Behara G, et al. Integrating metadata into deep autoencoder for handling prediction task of collaborative recommender system. Multim Tools Appl. 2024;83(14):42125–47

    2024

  81. [89]

    Metadata: an integral com- ponent of the modern data strategy

    Mohammed M, Talburt JR, Syed H. Metadata: an integral com- ponent of the modern data strategy. In: Congress in computer science, computer engineering, & applied computing (CSCE)

  82. [90]

    An empirical case study of meta-IP Chain DAO: the pioneer tokenless DAO

    Tang C, et al. An empirical case study of meta-IP Chain DAO: the pioneer tokenless DAO. In: IEEE 9th international confer - ence on data science in cyberspace (DSC). 2024. p. 24–31

    2024

  83. [91]

    Decentralised autonomous organizations (DAOs): an exploratory survey

    Tang C, et al. Decentralised autonomous organizations (DAOs): an exploratory survey. Distributed Ledger Technologies: Research and Practice, 2025. Human-Centric Intelligent Systems

    2025

  84. [92]

    Advancing continual lifelong learning in neural information retrieval: definition, dataset, framework, and empirical evaluation

    Hou J, Cosma G, Finke A. Advancing continual lifelong learning in neural information retrieval: definition, dataset, framework, and empirical evaluation. Inf Sci. 2025;687:121368

    2025

  85. [93]

    Brame: hierarchical data management framework for cloud-edge-device collaboration

    Liu X, et al. Brame: hierarchical data management framework for cloud-edge-device collaboration. 2025. https:// doi. org/ 10. 48550/ arXiv. 2502. 08331

    2025

  86. [94]

    On energy-aware and verifiable benchmarking of big data processing targeting AI pipe- lines

    Theodorou G, Karagiorgou S, Kotronis C. On energy-aware and verifiable benchmarking of big data processing targeting AI pipe- lines. In: IEEE international conference on Big Data (BigData)

  87. [95]

    Kubeedge

    Wang S, Hu Y, Wu J. Kubeedge. ai: Ai platform for edge devices

  88. [96]

    Multimodal archival data ecosystems

    Zhang Z, et al. Multimodal archival data ecosystems. In: IEEE international conference on web services (ICWS). 2024. p. 73–83

    2024

  89. [97]

    The future of multimodal artificial intelligence models for integrating imaging and clinical metadata: a narrative review

    Simon BD, et al. The future of multimodal artificial intelligence models for integrating imaging and clinical metadata: a narrative review. Diagnost Intervent Radiol. 2024

    2024

  90. [98]

    A generative AI-driven metadata modelling approach

    Bagchi M. A generative AI-driven metadata modelling approach

  91. [99]

    Leveraging retrieval augmented generative LLMs for automated metadata description generation to enhance data catalogs

    Singh M, et al. Leveraging retrieval augmented generative LLMs for automated metadata description generation to enhance data catalogs. 2025. https:// doi. org/ 10. 48550/ arXiv. 2503. 09003

    2025

  92. [100]

    Metadata creation and enrichment using artificial intelligence at meemoo

    Magnus B, et al. Metadata creation and enrichment using artificial intelligence at meemoo. J Digit Media Manag. 2025;13(2):110–23

    2025

  93. [101]

    Systematic literature review langchain proposed

    Asyrofi R, et al. Systematic literature review langchain proposed. In: International electronics symposium (IES). 2023. p. 533–7

    2023

  94. [102]

    Hugginggpt: solving AI tasks with chatgpt and its friends in hugging face

    Shen Y, et al. Hugginggpt: solving AI tasks with chatgpt and its friends in hugging face. Adv Neural Inf Process Syst. 2023;36:38154–80

    2023

  95. [103]

    TapeAgents: a holistic framework for agent development and optimization

    Bahdanau D, et al. TapeAgents: a holistic framework for agent development and optimization. 2024. https:// doi. org/ 10. 48550/ arXiv. 2412. 08445

    2024

  96. [104]

    Agentsquare: automatic llm agent search in mod- ular design space

    Shang Y, et al. Agentsquare: automatic llm agent search in mod- ular design space. 2024. https:// doi. org/ 10. 48550/ arXiv. 2410. 06153

    2024

  97. [105]

    TabPFN Unleashed: a scalable and effective solution to tabular classification problems

    Liu S-Y, Ye H-J. TabPFN Unleashed: a scalable and effective solution to tabular classification problems. 2025. https:// doi. org/

    2025

  98. [106]

    AI-powered policy management: implementing open policy agent (OPA) with intelligent agents in kubernetes

    Vadisetty R, Polamarasetti A. AI-powered policy management: implementing open policy agent (OPA) with intelligent agents in kubernetes. Cuestiones de Fisioterapia. 2025;54(5):19–27

    2025

  99. [107]

    AI explainability 360 toolkit

    Arya V, et al. AI explainability 360 toolkit. In: Proceedings of the 3rd ACM India joint international conference on data science & management of data. 2021. p. 376–9

    2021

  100. [108]

    48550/ arXiv. 2502. 02527

  101. [109]

    a study of blockchain-based metadata man- agement and its use for data verification

    Hori H, Oguchi M. a study of blockchain-based metadata man- agement and its use for data verification. In: Twelfth international symposium on computing and networking workshops (CAN- DARW). 2024. p. 63–8

    2024

  102. [110]

    A critical analysis of zero trust archi- tecture (ZTA)

    Fernandez EB, Brazhuk A. A critical analysis of zero trust archi- tecture (ZTA). Comput Stand Interfaces. 2024;89:103832

    2024

  103. [111]

    A standardized machine-readable dataset documen- tation format for responsible AI

    Jain N, et al. A standardized machine-readable dataset documen- tation format for responsible AI. 2024. https:// doi. org/ 10. 48550/ arXiv. 2407. 16883

    2024

  104. [112]

    The W3C data catalog vocabulary, ver - sion 2: rationale, design principles, and uptake

    Albertoni R, et al. The W3C data catalog vocabulary, ver - sion 2: rationale, design principles, and uptake. Data Intell. 2024;6(2):457–87

    2024

  105. [113]

    Toward total recall: enhancing FAIR- ness through AI-driven metadata standardization

    Sundaram SS, Musen MA. Toward total recall: enhancing FAIR- ness through AI-driven metadata standardization. 2025. https:// doi. org/ 10. 48550/ arXiv. 2504. 05307

    2025

  106. [114]

    FastMonitor: enhancing data access control with zero- trust architecture

    Mensah F. FastMonitor: enhancing data access control with zero- trust architecture. Int J Acad Indust Res Innov. 2024;10:347–51

    2024

  107. [115]

    Collaboration management for federated learn- ing

    Schlegel M, et al. Collaboration management for federated learn- ing. In: IEEE 40th international conference on data engineering workshops (ICDEW). 2024. p. 291–300

    2024

  108. [116]

    Amazon-KG: A knowledge graph enhanced cross- domain recommendation dataset

    Wang Y, et al. Amazon-KG: A knowledge graph enhanced cross- domain recommendation dataset. In: Proceedings of the 47th international ACM SIGIR conference on research and develop- ment in information retrieval. 2024. p. 123–30

    2024

  109. [117]

    Towards a metadata manage- ment system for provenance, reproducibility and accountabil - ity in federated machine learning

    Peregrina JA, Ortiz G, Zirpins C. Towards a metadata manage- ment system for provenance, reproducibility and accountabil - ity in federated machine learning. In: European conference on service-oriented and cloud computing. 2022. p. 5–18

    2022

  110. [118]

    FAIR assessment tools: evaluating use and per - formance

    Krans N, et al. FAIR assessment tools: evaluating use and per - formance. NanoImpact. 2022;27:100402

    2022

  111. [119]

    Metabench—a sparse benchmark to measure general ability in large language models

    Kipnis A, et al. Metabench—a sparse benchmark to measure general ability in large language models. 2024. https:// doi. org/

    2024

  112. [120]

    Graphql: a systematic mapping study

    Quiña-Mera A, et al. Graphql: a systematic mapping study. ACM Comput Surv. 2023;55(10):1–35

    2023

  113. [121]

    On the automated processing of user feedback

    Maalej W, et al. On the automated processing of user feedback. In: Handbook on natural language processing for requirements engineering. 2025, Springer. p. 279–308

    2025

  114. [122]

    Development of an information system with user-con- trolled structure and content

    Milev P. Development of an information system with user-con- trolled structure and content. Innov Inform Technol Econ Digital. 2024;1:7–12

    2024

  115. [123]

    48550/ arXiv. 2407. 12844

  116. [124]

    Continuous metadata in continuous integra - tion, stream processing and enterprise DataOps

    Underwood M. Continuous metadata in continuous integra - tion, stream processing and enterprise DataOps. Data Intell. 2023;5(1):275–88

    2023

  117. [127]

    Participatory approaches in AI develop- ment and governance: a principled approach

    Parthasarathy A, et al. Participatory approaches in AI develop- ment and governance: a principled approach. 2024. https:// doi. org/ 10. 48550/ arXiv. 2407. 13100. Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and insti...

    2024

  118. [2020]

    https:// doi. org/ 10. 48550/ arXiv. 2007. 09227

    2007

  119. [2024]

    https:// doi. org/ 10. 48550/ arXiv. 2501. 04008

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.