Pith. sign in

REVIEW 6 minor 75 references

SciSchema.org: A Multidisciplinary Collection of Schemas for Structured Scientific Process Descriptions

T0 review · 0 major / 6 minor · reviewed 2026-07-31 · grok-4.5

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

Pith's one-line read Sixteen expert-curated schemas turn prose-bound scientific methods into comparable, machine-checkable process descriptions across five domains.

desk verdict Solid first-release Data Descriptor: 16 dual-format process schemas with real HITL provenance; utility is prospective, not proven. read the letter →

arxiv 2607.27955 v1 pith:ZR7HR62Z submitted 2026-07-30 cs.DL cs.AIcs.CLcs.IR

Jennifer D'Souza , Sameer Sadruddin , Anisa Rula , Ana Bossler , Andrés Fullana , Enric Bas , Syed Ather , Defne Circi
show 18 more authors
Anlan Chen L. Catherine Brinson Alyssa Columbus George Demetriou Dongjun Jeong Tarun Kumar Frank Krüger Sascha Genehr Kai Budde-Sagert Anamaria Leonescu Francesco Lodola Chiara Florindi Gagana Balasubramanya Murthy Samson Oluwapelumi Olagbile Nazia Riasat Yan Sha Kevin Shen Shaokai Yang
This is my paper · ORCID
classification cs.DLcs.AIcs.CLcs.IR
keywords schemaminingscientificprocesslargelanguagemodelssymbolicknowledgestructuresgraphsJSONSHACLhuman-in-the-loop
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

Scientific methods are still mostly buried in papers—scattered across text, tables, figures, and supplements—so comparing studies, checking reporting gaps, or feeding methods into software and knowledge graphs is slow and error-prone. This paper releases SciSchema.org’s first collection: 16 expert-annotated schemas for concrete process types in biology and biotechnology, materials and chemistry, imaging and measurement, physics, and psychology. Each schema names the reusable fields needed to describe a process instance—inputs, materials, instruments or software, parameters, conditions, steps, measurements, outputs, and provenance—in both JSON Schema and SHACL form. The schemas were built with a three-stage human-in-the-loop workflow in which language models proposed structures from short process specs and batches of papers, experts corrected them, and experts then assembled final master schemas from the best candidates. The release also ships intermediate candidates, feedback records, source-paper metadata, and validation that the finals are well-formed and expert-reviewed, so others can annotate, extract, compare, and publish method-level records without reinventing the field list for each process.

What carries the argument

Human-in-the-loop scientific schema mining: language models generate candidate typed schemas from a short domain specification, then iteratively revise them against curated and expanded paper batches under periodic expert feedback; domain experts finally rate Stage-3 candidates and construct a master schema by selecting, editing, or merging preferred outputs.

What would settle it

Blinded extraction or annotation trials on held-out papers: measure agreement and completeness when humans or extractors populate the released master schemas versus fully hand-built gold schemas or versus raw Stage-3 model candidates, and check whether higher expert Likert scores predict higher population quality.

Watch

Extended reading notes

Core claim

The first SciSchema.org release is a documented, multidisciplinary set of 16 community expert-annotated scientific-process schemas, dual-encoded as JSON Schema and SHACL, produced by a staged human-in-the-loop language-model schema-mining workflow and checked for structure, provenance, expert preference, and syntactic conformance, and intended as reusable targets for annotation, metadata enrichment, knowledge graphs, information extraction, semantic publishing, and cross-study comparison.

Load-bearing premise

That experts merging model-proposed fields from roughly sixty papers per process produce field lists and nesting that are scientifically adequate community standards, even without a separate test of how well people or tools can fill those schemas from real papers.

Editorial extensions

If this is right

  • Method- and process-specific search can surface papers organized by shared attributes rather than full-text alone.
  • JSON Schema files can validate records, drive forms, and constrain extraction outputs; SHACL files can validate RDF process graphs and link into knowledge-graph templates.
  • Instantiated records can make unreported versus inapplicable fields visible, exposing reporting gaps without claiming to fix reproducibility.
  • The same field inventory supports cross-study comparison tables and machine-assisted or automated lab pipelines that need explicit inputs, parameters, and outputs.
  • Future community work can map these schemas onto domain vocabularies and repository profiles once the structural layer exists.

Reading between the lines

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

  • If the dual-format masters become default targets for extraction benchmarks, process-level F1 and gap-detection metrics could become as routine as entity extraction scores.
  • Shallow computational workflows were screened out because they yield few recurring fields; that selection rule may systematically under-represent pure software pipelines until schema depth criteria are relaxed.
  • Instruction-tuned models tended to expand more than reasoning-oriented ones in this workflow, suggesting extraction-target design may favor breadth-oriented generators when experts will prune later.
  • Without populated multi-paper corpora released alongside the schemas, adoption may lag until someone ships open annotated instances for at least one high-traffic process such as PCR or fatigue testing.
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

0 major / 6 minor

Summary. This Data Descriptor presents the first SciSchema.org release: 16 expert-annotated scientific-process schemas across Biology & Biotechnology, Materials & Chemistry, Imaging & Measurement, Physics, and Psychology. Each schema defines reusable fields (inputs, materials, instruments/software, parameters, conditions, steps, measurements, provenance) in dual JSON Schema and SHACL/Notation3 form. Schemas were produced via a documented three-stage human-in-the-loop workflow (schema-miner v3.2.5) in which twelve LLMs generated candidates from process specifications and paper corpora (~10 curated then ~50+ expanded papers), with periodic expert feedback, followed by expert construction of master schemas from Stage-3 candidates. The Zenodo archive includes master schemas, intermediate trajectories, feedback records, source-paper metadata, community materials, and analysis scripts. Technical validation covers input heterogeneity, feedback coverage, stage-wise growth in schema tokens/properties/leaves/depth, expert Likert ratings versus master-source selection (Spearman ρ=0.76), and full Draft 2020-12 / SHACL meta-validation.

Significance. If the resource is as documented, it supplies a concrete, reusable process-metadata layer that scholarly infrastructures largely lack, with dual machine-actionable formats suited to record validation, schema-guided IE, ORKG templates, and KG integration. Strengths include open dual-format masters, inspectable HITL provenance (stage outputs, feedback TSVs, metadata.csv), multi-model trajectories, syntactic conformance checks, and explicit prospective framing of reuse in Usage Notes. For cs.DL / scientific data venues, a multidisciplinary first release with community formation, selection criteria, and reproducible analysis scripts is a solid contribution even without external utility benchmarks.

minor comments (6)
  1. [Background & Summary; Usage Notes] Background & Usage Notes: State more explicitly that scientific adequacy rests on single-expert (or small-team) consolidation per process and that external utility tests (e.g., blinded IE F1, multi-annotator IAA on populated records, or comparison to fully manual gold schemas) are out of scope for this first release. This matches the paper’s actual claim but would reduce over-reading of “community standards.”
  2. [Methods (Master-schema construction and publication)] Methods, master-schema construction: Clarify how disagreements were handled when multiple experts contributed to one process (e.g., Catalytic Pyrolysis, Solvent Casting, Electrical Cell Stimulation) versus single-expert processes—consensus rule, lead editor, or merge protocol.
  3. [Data Overview; Figure 3] Figure 3 / Data Overview: Point readers to the exact tabular file in paper-analysis/ (or Zenodo path) that lists per-schema depth, total properties, and leaf counts so the scatter overview is fully reproducible from the text alone.
  4. [Technical Validation (Master expert-annotated schemas)] Technical Validation, Figure 9: Briefly note that the rating↔selection association is partly by design of the curation step (experts rate then choose/edit preferred candidates), so ρ=0.76 is consistency evidence rather than independent predictive validation.
  5. [Data Records; Table 1] Data Records: A one-line machine-readable process inventory (slug, domain, ORKG template ID, schema $id) alongside Table 1 would ease bulk reuse; Table 1 already helps but is prose-bound.
  6. [Methods (Process screening and selection)] Minor consistency: abstract and opening say “16 expert-annotated schemas” while Methods describe provisional acceptance of 18 with 16 confirming—fine, but a short clause on the two non-proceeding processes would close the loop.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: Data Descriptor releases constructed schemas via documented HITL workflow, not a derivation or prediction chain.

full rationale

SciSchema.org is a cs.DL Data Descriptor whose central claim is the release of 16 expert-annotated dual-format process schemas plus provenance materials, produced by a three-stage human-in-the-loop LLM schema-mining workflow and checked for structure, expert review, and JSON Schema/SHACL syntactic conformance. There is no first-principles derivation, fitted physical law, uniqueness theorem, or quantitative prediction whose output reduces to its inputs by construction. Self-citations to schema-miner [61, 62] document the open tool used to generate candidates; they do not smuggle an ansatz or forbid alternatives. Expert Likert ratings of Stage-3 candidates and subsequent master-schema selection (Figure 9, Spearman ρ=0.76) are the same curation act reported transparently, not an independent forecast presented as validation of a forced result. Usage Notes frame IE, forms, and KG integration as prospective capabilities. No circular step meets the quote-and-exhibit bar; score 0 with empty steps is the honest finding.

Assumptions & free parameters 3 free parameters · 5 assumptions · 2 invented entities

Load-bearing commitments are methodological and socio-technical rather than mathematical: literature maturity/granularity filters, adequacy of expert judgment over LLM candidates, and suitability of JSON Schema/SHACL as the interchange layer. No physical free parameters; ‘parameters’ are workflow design choices (batching, model set, Likert master construction).

free parameters (3)
  • Curated vs expanded corpus sizes (~10 and ~50+ papers per process) = ~10 curated; ~50+ expanded (varies by process)
    Hand-chosen corpus scales and batch splits drive what evidence enters schema-miner; not fit to a loss but chosen by protocol and availability.
  • Five-point Likert master-schema ratings and multi-source merge policy = 1–5 scale; 55.2% ratings ≥4 reported
    Final masters are selected/edited/combined using expert Likert judgments and free-text rules; different raters could yield different gold structures.
  • Process inclusion thresholds (>~15 years literature; high property granularity) = 18 provisionally accepted → 16 confirmed
    Screening criteria (informed by prior ALD >50-property experience) determine which of 29 proposals become the 16 schemas; changes the collection’s scope by design.
assumptions (5)
  • domain assumption Recurring process attributes reported across a mature literature can be abstracted into reusable schema fields that support comparison and IE.
    Stated throughout Background and Methods as the justification for schema mining from papers rather than only top-down standards.
  • domain assumption Domain experts can judge terminology, grouping, coverage, and scientific adequacy of candidate schemas well enough to define community master schemas.
    Master-schema construction and Likert protocol in Methods; no independent second-expert panel reported per process.
  • domain assumption JSON Schema Draft 2020-12 and SHACL/Notation3 can encode the same process-level content for document and KG ecosystems.
    Data Records and technical validation treat dual serialization as equivalent schema-level content.
  • ad hoc to paper Periodic expert feedback between paper batches is sufficient relative to per-paper feedback given modern LLM instruction following.
    Methods cites prior schema-miner studies for batching efficacy; adopted as operational axiom for this release’s cost control.
  • standard math Standard JSON parsing, Draft 2020-12 meta-schema validation, RDFLib parse, and pySHACL shape-graph checks establish syntactic release quality.
    Technical Validation section; conventional formalisms, not novel mathematics.
invented entities (2)
  • SciSchema.org schema collection (first release, 16 process schemas) independent evidence
    purpose: Named, versioned, community-maintained library of process-level metadata models with stable w3id/ORKG identifiers.
    The primary delivered artifact; existence is evidenced by the Zenodo deposit and schema catalog, but external uptake is not yet measured.
  • Per-process master schemas (e.g., CRISPR-Cas9, catalytic pyrolysis, neutrino event reconstruction)
    purpose: Gold-standard field inventories and nesting for describing process instances.
    Constructed by experts from model candidates; independent evidence is limited to expert annotation and syntactic validity, not third-party adoption studies in this paper.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SciSchema.org: A Multidisciplinary Collection of Schemas for Structured Scientific Process Descriptions." pith.science (2026). https://pith.science/paper/ZR7HR62Z

@misc{pith2026260727955,
  author       = {Pith},
  title        = {Pith review of: SciSchema.org: A Multidisciplinary Collection of Schemas for Structured Scientific Process Descriptions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZR7HR62Z}},
  note         = {Machine review of arXiv:2607.27955}
}
read the original abstract

Scientific processes are often described in heterogeneous article discourse, with details needed for comparison, reproducibility, reuse, and automation dispersed across prose, tables, figures, protocols, and supplementary files. We present the first release of SciSchema.org, a multidisciplinary collection of 16 expert-annotated schemas spanning Biology & Biotechnology, Materials & Chemistry, Imaging & Measurement, Physics, and Psychology. Each schema defines reusable fields for describing process instances, including inputs, outputs, materials, instruments or software, parameters, conditions, procedural steps, measurements, and provenance-related information. The schemas were created through a human-in-the-loop schema-mining workflow in which large language models generated candidate structures from process specifications, scientific articles, and expert feedback, followed by domain-expert construction of final master schemas. The dataset contains final schemas in JSON Schema and SHACL formats, intermediate model-generated schemas, expert-feedback records, source-paper metadata, community-development materials, and analysis scripts. Technical validation assessed schema structure, development provenance, expert review, and syntactic conformance. The collection supports structured annotation, metadata enrichment, scientific knowledge graphs, information extraction, semantic publishing, and cross-study comparison.

Figures

Figures reproduced from arXiv: 2607.27955 by the authors.

Figure 1
Figure 1. Panel 1. A broad query returns mostly full-text scholarly articles in search results. Panel 2a. A more method-specific query surfaces a schema-based structured comparison among the top five results. Panel 2b. A more process-specific query surfaces the same structured comparison at the top of the search results. Panel 3. Snapshot of the full structured comparison https://orkg.org/com parisons/R1471077, based on a sim… view at source ↗
Figure 2
Figure 2. The first SciSchema.org collection comprises 16 scientific process types organized into five broad disciplinary groups: Biology & Biotechnology, Materials & Chemistry, Imaging & Measurement, Physics, and Psychology. Each process type is associated with a schema intended to capture recurring inputs, conditions, instruments, computational steps, measurements, and outputs reported in the scien￾tific literature. schema-… view at source ↗
Figure 3
Figure 3. Structural overview of the released scientific-process schemas. [PITH_FULL_IMAGE:figures/full_fig_p014_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Average paper token length by process, aggregated across stage and batch groups. Error bars indicate the minimum and maximum paper token lengths. Domain experts reviewed the generated schemas after the designated stages and batches, and their written feedback was used …
Figure 5
Figure 5. Figure 5: Domain-expert feedback length across processes, models, and schema-miner [PITH_FULL_IMAGE:figures/full_fig_p016_5.png]
Figure 6
Figure 6. Figure 6: Schema length across processes, models, and workflow stages. [PITH_FULL_IMAGE:figures/full_fig_p017_6.png]
Figure 7
Figure 7. Figure 7: Process-level progression of schema structural complexity across stages and [PITH_FULL_IMAGE:figures/full_fig_p018_7.png]
Figure 8
Figure 8. Figure 8: Model-level progression of schema structural complexity across stages and [PITH_FULL_IMAGE:figures/full_fig_p019_8.png]
Figure 9
Figure 9. Figure 9: Relationship between final expert ratings and master-schema source selection. [PITH_FULL_IMAGE:figures/full_fig_p021_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

75 extracted references · 1 canonical work pages

  1. [1]

    The dataverse network®: An open-source application for sharing, dis- covering and preserving data.D-Lib Magazine17(2011)

    Magazine, D.-L. The dataverse network®: An open-source application for sharing, dis- covering and preserving data.D-Lib Magazine17(2011)

    2011

  2. [2]

    Datacite-a global registration agency for research data

    Brase, J. Datacite-a global registration agency for research data. In2009 fourth inter- national conference on cooperation and promotion of information resources in science and technology, 257–261 (IEEE, 2009). 25

    2009

  3. [3]

    & Feeney, P

    Hendricks, G., Tkaczyk, D., Lin, J. & Feeney, P. Crossref: The sustainable source of community-owned scholarly metadata.Quantitative Science Studies1, 414–427 (2020)

    2020

  4. [4]

    Knoth, P.et al.Core: A global aggregation service for open access papers.Scientific Data 10, 366 (2023)

    2023

  5. [5]

    & Orr, R

    Priem, J., Piwowar, H. & Orr, R. Openalex: A fully-open index of scholarly works, authors, venues, institutions, and concepts.arXiv preprint arXiv:2205.01833(2022)

    arXiv 2022

  6. [6]

    & Shotton, D

    Peroni, S. & Shotton, D. Opencitations, an infrastructure organization for open scholarship. Quantitative Science Studies1, 428–444 (2020)

    2020

  7. [7]

    & Smith, T

    Manghi, P., Bolikowski, L., Manola, N., Schirrwagen, J. & Smith, T. Openaireplus: the european scholarly communication data infrastructure.D-Lib Magazine18(2012)

    2012

  8. [8]

    Report, European Commission, Directorate-General for Research and Innovation, Luxembourg (2016)

    Ayris, P.et al.Realising the European Open Science Cloud: First Report and Recom- mendations of the Commission High Level Expert Group on the European Open Science Cloud. Report, European Commission, Directorate-General for Research and Innovation, Luxembourg (2016)

    2016

Show all 75 references
  1. [9]

    Arxiv at 20.Nature476, 145–147 (2011)

    Ginsparg, P. Arxiv at 20.Nature476, 145–147 (2011)

    2011

  2. [10]

    InHandbook on Neurosymbolic AI and Knowledge Graphs, 385–420 (SAGE Publications 1 Oliver’s Yard, 55 City Road, London, EC1Y 1SP, 2025)

    Auer, S.et al.Open research knowledge graph: a large-scale neuro-symbolic knowledge organization system. InHandbook on Neurosymbolic AI and Knowledge Graphs, 385–420 (SAGE Publications 1 Oliver’s Yard, 55 City Road, London, EC1Y 1SP, 2025)

    2025

  3. [11]

    D.et al.The fair guiding principles for scientific data management and stewardship.Scientific data3, 1–9 (2016)

    Wilkinson, M. D.et al.The fair guiding principles for scientific data management and stewardship.Scientific data3, 1–9 (2016)

    2016

  4. [12]

    M.et al.Shared metadata for data-centric materials science.Scientific Data10, 626 (2023)

    Ghiringhelli, L. M.et al.Shared metadata for data-centric materials science.Scientific Data10, 626 (2023)

    2023

  5. [13]

    Auer, S.et al.Improving access to scientific literature with knowledge graphs.Bibliothek Forschung und Praxis44, 516–529 (2020)

    2020

  6. [14]

    1,500 scientists lift the lid on reproducibility (2016)

    Baker, M. 1,500 scientists lift the lid on reproducibility (2016)

    2016

  7. [15]

    & Boutron, I

    Kapp, P., Esmail, L., Ghosn, L., Ravaud, P. & Boutron, I. Transparency and reporting characteristics of covid-19 randomized controlled trials.BMC medicine20, 363 (2022)

    2022

  8. [16]

    & Zare, R

    Metzger, N. & Zare, R. N. Interdisciplinary research: From belief to reality (1999)

    1999

  9. [17]

    The Human Genome Project (2025)

    National Human Genome Research Institute. The Human Genome Project (2025). URL https://www.genome.gov/human-genome-project. Last updated March 19, 2025; accessed July 1, 2026

    2025

  10. [18]

    Jumper, J.et al.Highly accurate protein structure prediction with alphafold.nature596, 583–589 (2021). 26

    2021

  11. [19]

    Tunyasuvunakool, K.et al.Highly accurate protein structure prediction for the human proteome.Nature596, 590–596 (2021)

    2021

  12. [20]

    The Higgs Boson: A Landmark Discovery (2012)

    ATLAS Experiment. The Higgs Boson: A Landmark Discovery (2012). URLhttps: //atlas.cern/Discover/Physics/Higgs. Higgs boson discovery announced July 4, 2012. Accessed July 2, 2026

    2012

  13. [21]

    P.et al.Observation of gravitational waves from a binary black hole merger

    Abbott, B. P.et al.Observation of gravitational waves from a binary black hole merger. Physical review letters116, 061102 (2016)

    2016

  14. [22]

    Eger, S.et al.Transforming science with large language models: A survey on ai-assisted scientific discovery, experimentation, content generation, and evaluation.arXiv preprint arXiv:2502.05151(2025)

    2025

  15. [23]

    Zhang, Y.et al.Exploring the role of large language models in the scientific method: from hypothesis to discovery.npj Artificial Intelligence1, 14 (2025)

    2025

  16. [24]

    InProceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, 17744–17761 (2025)

    Zheng, T.et al.From automation to autonomy: A survey on large language models in scientific discovery. InProceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, 17744–17761 (2025)

    2025

  17. [25]

    Hillson, N.et al.Building a global alliance of biofoundries.Nature communications10, 2040 (2019)

  18. [26]

    Mehr, S. H. M., Craven, M., Leonov, A. I., Keenan, G. & Cronin, L. A universal system for digitization and automatic execution of the chemical synthesis literature.Science370, 101–108 (2020)

    2020

  19. [27]

    A., MacKnight, R., Kline, B

    Boiko, D. A., MacKnight, R., Kline, B. & Gomes, G. Autonomous chemical research with large language models.Nature624, 570–578 (2023)

    2023

  20. [28]

    Tom, G.et al.Self-driving laboratories for chemistry and materials science.Chemical Reviews124, 9633–9732 (2024)

    2024

  21. [29]

    B.et al.Science acceleration and accessibility with self-driving labs.Nature Communications16, 3856 (2025)

    Canty, R. B.et al.Science acceleration and accessibility with self-driving labs.Nature Communications16, 3856 (2025)

    2025

  22. [30]

    UNESCO Recommendation on Open Science.https://www.unesco.org/en/ open-science/about(2023)

    UNESCO. UNESCO Recommendation on Open Science.https://www.unesco.org/en/ open-science/about(2023). Last updated 21 September 2023; accessed 3 July 2026

    2023

  23. [31]

    Recommendation of the council concerning access to research data from public funding

    OECD. Recommendation of the council concerning access to research data from public funding. OECD Legal Instruments, OECD/LEGAL/0347 (2021). URLhttps://legali nstruments.oecd.org/en/instruments/OECD-LEGAL-0347. Adopted 14 December 2006; amended 20 January 2021; accessed 3 July 2026

    2021

  24. [32]

    Ensuring free, immediate, and equitable access to federally funded research

    Nelson, A. Ensuring free, immediate, and equitable access to federally funded research. Memorandum for the Heads of Executive Departments and Agencies (2022). URLhttps: //bidenwhitehouse.archives.gov/wp-content/uploads/2022/08/08-2022-OSTP-Pub lic-Access-Memo.pdf. Accessed 3 J...

    2022

  25. [33]

    Data management & sharing policy overview

    National Institutes of Health. Data management & sharing policy overview. NIH Grants & Funding (2023). URLhttps://grants.nih.gov/policy-and-compliance/policy -topics/sharing-policies/dms/policy-overview. Policy effective 25 January 2023; accessed 3 July 2026

    2023

  26. [34]

    State council releases regulation on scientific data management

    State Council of the People’s Republic of China. State council releases regulation on scientific data management. English.gov.cn; Xinhua (2018). URLhttps://english.ww w.gov.cn/policies/latest_releases/2018/04/02/content_281476099479814.htm. Accessed 3 July 2026

    2018

  27. [35]

    National policy on promoting open access to publicly funded scholarly publications and scientific data

    Integrated Innovation Strategy Promotion Council. National policy on promoting open access to publicly funded scholarly publications and scientific data. Provisional translation, Cabinet Office, Government of Japan (2024). URLhttps://www8.cao.go.jp/cstp/oa_2 40216_en.pdf. Acce...

    2024

  28. [36]

    JST policy on open access to research publications and research data management

    Japan Science and Technology Agency. JST policy on open access to research publications and research data management. Provisional translation (2025). URLhttps://www.jst. go.jp/EN/about/openscience/policy_openscience_en_r7.pdf. Originally issued 1 April 2017; revised 1 April 20...

    2025

  29. [37]

    Tri-agency research data management policy

    Canadian Institutes of Health Research, Natural Sciences and Engineering Research Coun- cil of Canada & Social Sciences and Humanities Research Council of Canada. Tri-agency research data management policy. Government of Canada, Science.gc.ca (2021). URL https://science.gc.ca/...

    2021

  30. [38]

    NHMRC and MRFF open science policy

    National Health and Medical Research Council & Medical Research Future Fund. NHMRC and MRFF open science policy. National Health and Medical Research Council (2025). URLhttps://www.nhmrc.gov.au/about-us/resources/nhmrc-and-mrff-open-scien ce-policy. Accessed 3 July 2026

    2025

  31. [39]

    A european strategy for data

    European Commission. A european strategy for data. Shaping Europe’s Digital Future (2026). URLhttps://digital-strategy.ec.europa.eu/en/policies/strategy-data. Last updated 22 June 2026; accessed 3 July 2026

    2026

  32. [40]

    Regulation (EU) 2022/868 of the european parliament and of the council of 30 may 2022 on european data governance and amending regulation (EU) 2018/1724 (Data Governance Act)

    European Parliament and Council of the European Union. Regulation (EU) 2022/868 of the european parliament and of the council of 30 may 2022 on european data governance and amending regulation (EU) 2018/1724 (Data Governance Act). Official Journal of the European Union, L 152,...

    2022

  33. [41]

    European Parliament and Council of the European Union. Regulation (EU) 2023/2854 of the european parliament and of the council of 13 december 2023 on harmonised rules on fair access to and use of data and amending regulation (EU) 2017/2394 and directive (EU) 2020/1828 (Data Ac...

    2023

  34. [42]

    European Parliament and Council of the European Union. Regulation (EU) 2024/903 of the european parliament and of the council of 13 march 2024 laying down measures for a high level of public sector interoperability across the union (Interoperable Europe Act). Official Journal ...

    2024

  35. [43]

    V., Brickley, D

    Guha, R. V., Brickley, D. & Macbeth, S. Schema.org: evolution of structured data on the web.Communications of the ACM59, 44–51 (2016)

    2016

  36. [44]

    Gray, A. J. G., Castro, L. J., Juty, N. & Goble, C. Schema.org for scientific data. In Artificial Intelligence for Science, chap. 27, 495–514 (World Scientific, 2023)

    2023

  37. [45]

    J., Goble, C

    Gray, A. J., Goble, C. A., Jimenez, R. C.et al.Bioschemas: From potato salad to protein annotation.ISWC (posters, demos & industry tracks)2017(2017)

    2017

  38. [46]

    Bioschemas: Improving life sciences discoverability on the web

    Bioschemas Community. Bioschemas: Improving life sciences discoverability on the web. Online (2015). Accessed: 2025-06-12. Available: https://bioschemas.org/

    2015

  39. [47]

    Nishio, K.et al.A digital laboratory with a modular measurement system and standardized data format.Digital Discovery4, 1734–1742 (2025)

    2025

  40. [48]

    & Velterop, J

    Groth, P., Gibson, A. & Velterop, J. The anatomy of a nanopublication.Information services and use30, 51–56 (2010)

    2010

  41. [49]

    In2018 IEEE 14th International Conference on e-Science (e-Science), 83–92 (IEEE, 2018)

    Kuhn, T.et al.Nanopublications: a growing resource of provenance-centric scientific linked data. In2018 IEEE 14th International Conference on e-Science (e-Science), 83–92 (IEEE, 2018)

    2018

  42. [50]

    Ashburner, M.et al.Gene ontology: tool for the unification of biology.Nature genetics 25, 25–29 (2000)

    2000

  43. [51]

    & Hastings, J

    Neuhaus, F. & Hastings, J. Ontology development is consensus creation, not (merely) representation.Applied Ontology17, 495–513 (2022)

    2022

  44. [52]

    Matentzoglu, N.et al.Ontology development kit: a toolkit for building, maintaining and standardizing biomedical ontologies.Database2022, baac087 (2022)

    2022

  45. [53]

    Amershi, S., Cakmak, M., Knox, W. B. & Kulesza, T. Power to the people: The role of humans in interactive machine learning.AI magazine35, 105–120 (2014)

    2014

  46. [54]

    L., Hogan, A

    Martinez-Rodriguez, J. L., Hogan, A. & Lopez-Arevalo, I. Information extraction meets the semantic web: a survey.Semantic Web11, 255–335 (2020)

    2020

  47. [55]

    & Auer, S

    Shamsabadi, M., D’Souza, J. & Auer, S. Large language models for scientific information extraction: An empirical study for virology. In Graham, Y. & Purver, M. (eds.)Findings 29 of the Association for Computational Linguistics: EACL 2024, 374–392 (Association for Computational...

    2024

  48. [56]

    Dagdelen, J.et al.Structured information extraction from scientific text with large lan- guage models.Nature communications15, 1418 (2024)

    2024

  49. [57]

    D’Souza, J.et al.Mining for species, locations, habitats, and ecosystems from scientific papers in invasion biology: A large-scale exploratory study with large language models. In Basile, V.et al.(eds.)Proceedings of the 1st Workshop on Ecology, Environment, and Nat- ural Lang...

    2025

  50. [58]

    Mior, M. J. Large language models for json schema discovery.arXiv preprint arXiv:2407.03286(2024)

    2024 arXiv

  51. [59]

    & Simperl, E

    Zhang, B., He, Y., Pintscher, L., Mero˜ no-Pe˜ nuela, A. & Simperl, E. Schema genera- tion for large knowledge graphs using large language models. In Christodoulopoulos, C., Chakraborty, T., Rose, C. & Peng, V. (eds.)Findings of the Association for Computa- tional Linguistics:...

    2025

  52. [60]

    & Paton, N

    Wu, Z., Chen, J. & Paton, N. W. Schema inference for tabular data repositories using large language models.arXiv preprint arXiv:2509.04632(2025)

    2025 arXiv

  53. [61]

    InEuropean Semantic Web Conference, 244– 261 (Springer, 2025)

    Sadruddin, S.et al.Llms4schemadiscovery: A human-in-the-loop workflow for scientific schema mining with large language models. InEuropean Semantic Web Conference, 244– 261 (Springer, 2025)

    2025

  54. [62]

    Sadruddin, S.et al.Schema-miner pro: Agentic ai for ontology grounding over llm- discovered scientific schemas in a human-in-the-loop workflow.Semantic Web17(2026)

    2026

  55. [63]

    George, S. M. Atomic layer deposition: an overview.Chemical reviews110, 111–131 (2010)

    2010

  56. [64]

    R., Goulas, A

    van Ommen, J. R., Goulas, A. & Puurunen, R. L. Atomic layer deposition.Kirk-Othmer Encyclopedia of Chemical Technology1–42 (2021)

    2021

  57. [65]

    Kessels, E.et al.Atomic layer deposition.Nature Reviews Methods Primers5, 66 (2025)

    2025

  58. [66]

    W.et al.Scaling instruction-finetuned language models.Journal of Machine Learning Research25, 1–53 (2024)

    Chung, H. W.et al.Scaling instruction-finetuned language models.Journal of Machine Learning Research25, 1–53 (2024)

    2024

  59. [67]

    Guo, D.et al.Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning.arXiv preprint arXiv:2501.12948(2025)

    2025 arXiv

  60. [68]

    Hoffmann, J.et al.Training compute-optimal large language models.arXiv preprint arXiv:2203.1555610(2022). 30

    2022 arXiv

  61. [69]

    H.et al.Ministral 3.arXiv preprint arXiv:2601.08584(2026)

    Liu, A. H.et al.Ministral 3.arXiv preprint arXiv:2601.08584(2026)

    2026 arXiv

  62. [70]

    Qwen3 technical report (2025)

    Team, Q. Qwen3 technical report (2025). URLhttps://arxiv.org/abs/2505.09388. 2505.09388

    2025 arXiv

  63. [71]

    Gemma 3 (2025)

    Team, G. Gemma 3 (2025). URLhttps://goo.gle/Gemma3Report

    2025

  64. [72]

    Zenodo (2026)

    D’Souza, J.et al.SciSchema.org: First Release Dataset. Zenodo (2026). URLhttps: //doi.org/10.5281/zenodo.21254268. Dataset

    2026 doi

  65. [73]

    URLhttps://github.com/RDFLib/rdflib

    Krech, D.et al.RDFLib (2026). URLhttps://github.com/RDFLib/rdflib. Software

    2026

  66. [74]

    & Car, N

    Sommer, A. & Car, N. pySHACL (2026). URLhttps://github.com/RDFLib/pySHACL. Software, Apache License 2.0

    2026

  67. [75]

    & Kunkel, J

    Doosthosseini, A., Decker, J., Nolte, H. & Kunkel, J. Saia: A seamless slurm-native solution for hpc-based services.The Journal of Supercomputing82, 403 (2026). 31

    2026

Pith tools

Reviewed July 31, 2026 · model on record in the stance chip above.