1. Policy Statement and Scientific Purpose

Neutrosophic Optimization and Intelligent Systems (NOIS), published by Sciences Force LLC, is committed to the transparency, reliability, verifiability, and responsible reuse of scholarly research.

The journal recognizes that scientific claims must be supported by evidence appropriate to the nature of the research. In mathematical studies, this may require precise definitions, valid proofs, and transparent assumptions. In optimization and computational research, it may require executable algorithms, documented experimental conditions, reproducible numerical results, and fair comparisons. In artificial intelligence research, it may additionally require traceable datasets, transparent evaluation protocols, and adequate documentation of models and training procedures.

This policy establishes the requirements for research data, software, computational workflows, mathematical verification, reporting transparency, availability statements, and responsible sharing of supporting research materials.

It applies to original research, methodological studies, computational investigations, simulation-based studies, systematic reviews, and other scholarly contributions where supporting data or research materials are relevant.

Requirements are applied proportionately to the type of contribution. Purely theoretical research is not required to create datasets or software that were not used, while computational or empirical research must provide sufficient evidence to support its reported findings.

The policy is informed by the Committee on Publication Ethics (COPE), the Transparency and Openness Promotion Guidelines (TOP 2025), the FAIR Guiding Principles for scientific data management, and the publisher's Data, Code, Materials and Reproducibility framework.

Its central objective is to make research claims independently assessable without compromising legitimate privacy, security, intellectual property, or ethical obligations.

2. Research Transparency and Reproducibility

NOIS distinguishes among several related aspects of scientific verification.

Transparency concerns the accurate disclosure of the research question, assumptions, methods, materials, analytical decisions, limitations, and evidence used to support a claim.

Computational reproducibility concerns the ability to obtain the reported computational findings using the documented data, code, software, and analytical procedures under appropriately comparable conditions.

Replicability concerns the ability of an independent study to obtain scientifically consistent findings through a new implementation, independent execution, or appropriately comparable research design.

Mathematical verification concerns the independent examination of definitions, logical arguments, assumptions, proofs, derivations, and theoretical conclusions.

Different studies may require different verification methods. Exact numerical agreement is not always expected for nondeterministic computations, but authors must explain relevant sources of variation and establish appropriate criteria for evaluating consistency.

A research claim must not be described as experimentally validated, independently reproduced, or mathematically proven unless the evidence justifies that description.

3. General Requirements for All Research Submissions

All research submissions must describe their methods and supporting evidence with sufficient clarity for meaningful editorial and peer assessment.

Authors must provide a Data Availability Statement, including when no new data were generated or analyzed.

Where computational code, software, or analytical workflows are material to the findings, authors must also provide a Code Availability Statement.

Empirical and computational studies must identify the relevant data sources, methodological procedures, software environments, experimental configurations, and essential analytical decisions.

Where research protocols, preregistrations, analysis plans, or specialist reporting guidelines apply, authors must disclose the relevant information and explain material deviations.

Authors must distinguish original contributions from established methods and accurately cite data, software, models, and research materials developed by others.

Information supplied at submission must agree with the manuscript, supporting files, repository records, and subsequent publication metadata.

The journal may return submissions for clarification or technical completion where essential information is missing.

4. Mandatory Data Availability Statement

Every research submission must include a clearly labeled Data Availability subsection within its Statements and Declarations.

The statement must accurately describe the supporting data relevant to the publication.

Depending on the study, it must identify:

- The existence and nature of the data.

- Whether data were generated, collected, simulated, reused, or analyzed.

- The repository or other location where data are available.

- Relevant persistent identifiers and data versions.

- Conditions governing access and reuse.

- Justified restrictions on disclosure.

- A legitimate procedure for requesting restricted data where feasible.

- Whether no new data were generated or analyzed.

A general statement that data are available must not be used when the material cannot actually be obtained through the stated route.

Where several datasets have different access conditions, the statement must distinguish them.

Where data are deposited under an embargo, the statement must identify the reason and relevant access arrangements.

For studies that do not generate or analyze data, an accurate declaration of non-applicability is acceptable.

Appropriate Statement for Purely Theoretical Research

The following wording may be used when factually accurate:

Data Availability: No new data were generated or analyzed in this theoretical study.

Appropriate Statement for Data Included in an Article

The following wording may be used when factually accurate:

Data Availability: All data supporting the reported findings are provided within the article and its supplementary materials.

Other circumstances require a study-specific statement accurately identifying the actual repository, dataset, access conditions, and restrictions.

Authors must not copy a standard declaration that misrepresents their research.

5. Code and Software Availability

A Code Availability Statement is required when custom software, scripts, algorithms, simulation programs, or computational workflows contribute materially to the reported findings.

Authors must identify whether the relevant code is openly available, accessible under stated conditions, restricted for legitimate reasons, or unavailable.

Where code is openly shared, the statement should identify the repository, preserved version or release, persistent identifier where available, and applicable license.

Where sharing is restricted, the statement must explain the relevant legal, ethical, security, contractual, or intellectual property limitations.

The journal expects that essential computational methods will be documented sufficiently for meaningful independent evaluation, even where the original implementation cannot be publicly distributed.

Code availability is not applicable to a purely analytical manuscript that does not develop or employ computational code.

The use of widely available commercial software does not necessarily require distribution of its proprietary source code. Authors must nevertheless identify the software and provide sufficient details of their use of it.

A link to a software repository does not, by itself, establish that the results are reproducible.

6. Research Repositories and Persistent Identification

Authors are encouraged to deposit shareable data, code, and research materials in trusted disciplinary, institutional, or general-purpose repositories.

Suitable repositories should support persistent identification, meaningful metadata, clear access conditions, and appropriate long-term availability.

Research objects should be assigned persistent identifiers, such as DOIs, where the chosen repository supports them.

Authors should prefer repository records that remain discoverable even when access to the underlying files is restricted.

Temporary file-sharing links, unversioned personal websites, and short-lived downloads are not preferred as the sole means of distributing essential research materials.

Repositories are not interchangeable in their technical features, storage guarantees, preservation arrangements, and licensing conditions. Authors must select services appropriate to the materials and applicable restrictions.

The journal does not require the exclusive use of any single commercial repository.

7. FAIR Data Principles

NOIS supports the FAIR Principles: Findability, Accessibility, Interoperability, and Reusability.

Findability: Research objects should have meaningful descriptions, searchable metadata, and persistent identifiers where available.

Accessibility: Access methods, permissions, restrictions, and contact procedures must be stated clearly.

Interoperability: Data should use appropriate formats, definitions, standards, and documentation that facilitate interpretation and integration.

Reusability: Research objects should include adequate provenance, methods, contextual information, and licensing terms to support legitimate further analysis.

FAIR compliance does not require unrestricted public disclosure of confidential or sensitive information.

Restricted datasets can follow FAIR principles through publicly discoverable metadata and controlled access arrangements.

Authors should adopt the degree of technical implementation appropriate to their research domain and the resources involved.

8. Data Provenance, Documentation, and Processing

Authors must document the origins and transformations of data used to support their findings.

Where relevant, documentation should include:

- The data source and original collection or acquisition method.

- The date or version of acquisition.

- Inclusion and exclusion criteria.

- Variable definitions, units, labels, and coding conventions.

- Data cleaning and quality-control procedures.

- Missing-data identification and treatment.

- Normalization, scaling, transformation, or feature construction.

- Removal or exclusion of observations and the reasons for doing so.

- Aggregation and sampling procedures.

- Data partitioning and changes between processing stages.

- The relationship between source data, analyzed data, and published results.

Authors should make clear whether reported findings are based on original records, transformed data, selected subsets, or derived measurements.

Material transformations must not be concealed or represented as original observations.

Where appropriate, authors should provide a README, data dictionary, processing scripts, or equivalent documentation.

9. Third-Party, Licensed, and Dynamic Datasets

Authors using previously published or externally obtained datasets must identify their sources and comply with relevant licensing and access conditions.

The original data creators and relevant publications must be acknowledged and cited.

Authors must not redistribute third-party data where the applicable license or agreement prohibits redistribution.

Where data cannot be redistributed, the manuscript should identify how qualified researchers may obtain the source material independently, subject to applicable restrictions.

For dynamic databases, continuously updated collections, streaming data, or changing online resources, authors must identify the version, snapshot, retrieval date, query parameters, and selection criteria used in the study.

Where feasible, an immutable snapshot or a reproducible extraction specification should be preserved.

Authors must distinguish between the dataset actually used and later versions of the same resource.

A link to a changing database without adequate version information may be insufficient for computational verification.

10. Synthetic Data and Simulation Transparency

Studies using simulated, synthetic, or artificially generated data must identify the data as such.

Authors must report the process used to generate the data, including relevant mathematical assumptions, distributions, constraints, parameter values, and generation algorithms.

Where a generative model or another artificial intelligence system produces synthetic data, the manuscript must identify the relevant model, its role, important settings, and limitations.

Authors must distinguish synthetic data from real-world measurements, human observations, clinical records, and independently collected empirical evidence.

Synthetic datasets must not be presented as evidence of real-world performance unless the associated validation justifies that interpretation.

Simulation studies must describe the number of scenarios, experimental repetitions, parameter settings, initialization procedures, and methods of analysis.

Where randomness affects the results, reproducible random-number generation procedures and seeds should be provided whenever feasible.

Authors must explain how the simulation design reflects the research question and acknowledge limitations concerning generalizability.

11. Mathematical Proofs and Theoretical Verifiability

For mathematical and theoretical research, scientific reproducibility primarily concerns the validity and independent verification of the underlying reasoning.

Manuscripts must define all essential mathematical objects, assumptions, notation, operators, constraints, and domains of application.

Theorems, lemmas, propositions, and corollaries must be stated precisely and supported by logically valid arguments.

Claims of existence, uniqueness, convergence, optimality, consistency, stability, equivalence, or generality must identify the conditions under which they hold.

Authors must distinguish exact mathematical proofs from numerical evidence, examples, conjectures, and computer-assisted verification.

Where an argument depends on an earlier theorem, the relevant result and its assumptions must be cited accurately.

Computational demonstrations cannot replace proof of a general theorem unless the methodology provides a mathematically justified exhaustive or certified verification.

If a computer-assisted proof is used, authors must describe the verification procedure and provide relevant implementation details, input conditions, and supporting evidence.

Where a result is valid only under additional restrictions, those restrictions must be reported rather than omitted to strengthen the claim.

12. Neutrosophic and Uncertainty-Modeling Research

Research introducing or modifying neutrosophic, fuzzy, interval-valued, or other uncertainty-related structures must present a precise mathematical foundation.

Authors must clearly define the form and interpretation of all uncertainty components, including truth, indeterminacy, and falsity degrees where relevant.

The admissible domains, mathematical constraints, aggregation rules, and other conditions must be explicitly stated.

Assumptions must be appropriate to the specific neutrosophic or uncertainty framework. Constraints that apply to one representation must not be imposed on a different representation without mathematical justification.

When proposing a new model or operator, authors should examine relevant properties such as consistency, monotonicity, idempotency, boundary behavior, stability, or other properties that are material to the proposed contribution.

The manuscript should explain how the proposed framework relates to established methods and whether the new results reduce to known structures under special conditions.

Illustrative examples must be internally consistent and must use the definitions actually presented.

Where a method includes numerical calculations, authors must provide enough information to verify those calculations independently.

A change in terminology, symbolic notation, or parameter naming alone does not establish a new mathematical contribution.

13. Optimization Algorithms and Implementation Transparency

Research proposing or evaluating optimization algorithms must describe the method with sufficient precision for independent implementation or verification.

Where applicable, the manuscript must identify:

- The optimization problem and objective functions.

- Decision variables and their permissible domains.

- Constraints and feasibility conditions.

- Initialization procedures.

- Update equations and algorithmic operators.

- Control parameters and their chosen values.

- Constraint-handling mechanisms.

- Selection, replacement, and acceptance rules.

- Stopping conditions and computational budgets.

- Randomization procedures.

- Output definitions and performance criteria.

Pseudocode should agree with the mathematical description and the implementation used for experiments.

If the proposed method modifies an existing algorithm, authors must explain the changes precisely and justify their significance.

Where convergence, computational complexity, optimality, or performance guarantees are claimed, appropriate theoretical or empirical support must be provided.

Source code or executable workflows should be made available where legally and ethically permissible.

If the implementation cannot be shared openly, the manuscript must explain the restriction and provide sufficient methodological detail for independent assessment.

14. Benchmarking and Fair Algorithm Comparison

Comparative optimization studies must use relevant benchmark problems and baseline algorithms appropriate to the research claims.

Authors must identify the benchmark sources, versions, dimensions, constraints, and relevant problem characteristics.

Computational comparisons must be conducted under conditions that are sufficiently comparable to support the conclusions drawn.

Where relevant, these conditions include:

- Evaluation budgets and termination criteria.

- Numbers of independent runs.

- Initialization and randomization procedures.

- Parameter tuning protocols.

- Constraint-handling methods.

- Hardware and software environments.

- Time limits and computational resource budgets.

- Definitions of reported performance measures.

The use of identical iteration counts does not necessarily establish a fair comparison when methods differ substantially in the number or cost of objective-function evaluations.

Authors must justify the computational budget or comparison metric used.

Performance summaries should reflect variability across runs and should not rely exclusively on the most favorable observed result.

Claims of superior performance must be supported by appropriate evidence, which may include statistical comparisons, effect sizes, uncertainty measures, scalability experiments, or convergence analysis.

Negative results, performance trade-offs, and cases where the proposed method is inferior to established approaches must not be selectively concealed.

15. Artificial Intelligence and Machine Learning Reproducibility

Manuscripts involving machine learning, deep learning, intelligent systems, or related artificial intelligence methods must provide adequate information about the data, models, training procedures, and evaluation design.

Depending on the study, authors must identify:

- Data sources, provenance, characteristics, and relevant licenses.

- Sample sizes, feature definitions, and preprocessing procedures.

- Training, validation, and test dataset partitions.

- The basis for choosing cross-validation or other evaluation strategies.

- Model architectures, learning procedures, and hyperparameters.

- Pretrained models, checkpoints, or external services used.

- Sources of randomness and initialization settings.

- Training duration and computational resources.

- Evaluation metrics and their definitions.

- Baseline models and comparison protocols.

- Procedures used to prevent training-test leakage.

- Potential dataset bias and limitations of generalizability.

When the same records, subjects, institutions, or related observations could occur in multiple partitions, authors must explain how relevant leakage risks were managed.

Hyperparameter selection and model comparison should avoid inappropriate reuse of test data.

Where class imbalance, unequal error consequences, calibration, fairness, or domain shift are material, appropriate evaluation must be provided.

Authors should share code, configurations, trained-model information, and evaluation scripts where lawful and feasible.

Claims of real-world effectiveness must not exceed the evidence provided by the study.

16. Reproducible Computational Environments

Computational research must identify the technical environment relevant to obtaining the reported findings.

Documentation should include, where material:

- Programming languages and versions.

- Operating system and version.

- Relevant software packages, libraries, and dependencies.

- Numerical solvers and settings.

- Hardware characteristics, including CPUs, GPUs, and memory where relevant.

- Random seeds and random-number generators.

- Numerical precision and tolerance settings.

- Parallel or distributed execution arrangements.

- Data preprocessing and model-training configurations.

- Commands or scripts required to reproduce principal outputs.

Authors are encouraged to provide environment specifications, dependency lockfiles, container definitions, or other reproducible configuration files.

A README should explain the intended execution order, input requirements, installation procedures, and expected outputs.

Where computations require substantial resources, authors should identify approximate resource needs and provide a reduced test case where feasible.

When bit-for-bit reproduction is not realistic because of nondeterministic hardware, parallel computing, external services, or numerical variability, authors must explain the expected variation and the appropriate verification criteria.

17. Generative AI and External Model Services

When generative artificial intelligence is used as part of research methodology, authors must report its role transparently.

Material methodological details may include the model or service name, available version information, access date, important system settings, relevant prompt templates, retrieval configurations, input preprocessing, output selection, and human post-processing.

Where the study depends on repeated interactions with a generative model, authors should explain the evaluation protocol, sampling strategy, and variability of outputs.

The use of a commercial model through an external API may limit exact reproduction because model versions, service behavior, or terms of access can change.

Such limitations must be reported honestly.

Where disclosure of particular prompts, datasets, weights, or configurations is restricted, authors must explain the reason and supply an appropriate alternative form of verification where possible.

AI-generated synthetic observations must be labeled as such.

Authors must not claim that a model has independently confirmed a mathematical theorem, validated a dataset, or established an empirical finding unless an appropriate verification procedure has actually been conducted.

Human authors remain accountable for the reliability and interpretation of AI-assisted results.

18. Statistical Analysis and Reporting

Quantitative studies must use analytical methods appropriate to their research design, data structure, and scientific questions.

Authors must describe the relevant statistical tests, assumptions, significance criteria, uncertainty measures, and procedures for handling multiple comparisons where applicable.

The number of observations, independent experimental units, and repeated measurements must be reported accurately.

Summary measures must be appropriate to the data and should not conceal material variability.

Where inferential statistics are used, authors should report relevant estimates and measures of uncertainty, such as confidence intervals, alongside p-values where suitable.

Authors must distinguish statistical significance from practical or scientific importance.

Exploratory analyses must not be presented as confirmatory analyses without appropriate justification.

Where assumptions are violated, the limitations and chosen remedies must be explained.

Computational tables and figures must agree with the underlying analyses.

Selective reporting of favorable outcomes, unjustified exclusion of observations, and manipulation of statistical results are unacceptable.

19. Research Protocols, Preregistration, and Reporting Standards

Where discipline-specific reporting guidelines or research registration standards apply, authors must follow the applicable requirements or justify relevant deviations.

Studies involving prospective clinical trials or other regulated interventional research must comply with the journal's applicable ethical and registration requirements.

For preregistered studies, authors must identify the relevant registration record, protocol, or analysis plan and distinguish planned analyses from subsequent exploratory analyses.

Material deviations from preregistered methods must be disclosed and justified.

Systematic reviews must report the methods used to identify, select, assess, and synthesize the relevant literature.

Computational and simulation studies should provide adequate methodological documentation even when formal preregistration is not customary in the discipline.

Purely theoretical mathematical research is not required to register a clinical trial or provide an empirical study checklist.

Reporting requirements must be matched to the actual design and contribution.

20. Sensitive, Restricted, and High-Risk Data

NOIS supports responsible data sharing and does not require authors to disclose information where doing so would violate applicable law, research ethics, privacy obligations, security requirements, or binding contractual restrictions.

Legitimate restrictions may concern identifiable personal information, confidential institutional records, protected commercial information, restricted third-party datasets, or materials whose release would create a substantial risk of harm.

Authors must disclose the nature of the restriction and provide a lawful means of independent assessment where feasible.

Appropriate alternatives may include:

- Controlled-access repositories.

- Data-use agreements.

- Secure research environments.

- Independent verification by authorized specialists.

- Publicly available aggregate statistics.

- Carefully generated synthetic examples.

- Redacted methods or documentation.

- Reproducible scripts that operate on an appropriately shareable example dataset.

Synthetic or redacted materials must not be represented as identical to the original restricted data.

The journal may seek specialist ethical or legal advice where access restrictions materially affect the assessment of a submission.

The absence of public access to sensitive data is not, by itself, evidence of misconduct.

21. Research Data Integrity and Original Records

Authors must retain relevant original research records in accordance with applicable institutional, disciplinary, contractual, and legal requirements.

Original records may include observations, source datasets, computational logs, experimental outputs, image files, code versions, simulation results, and other evidence material to the findings.

Authors must not fabricate, falsify, selectively manipulate, or misrepresent research data.

Changes to source information must be accurately documented when they are material to scientific interpretation.

Where data are transformed or cleaned, authors should preserve sufficient provenance information to explain the relationship between original and processed records.

Errors must be corrected transparently and should not be concealed by replacing records without appropriate documentation.

The journal may request relevant original evidence during review or post-publication assessment.

The response to any inability to provide original materials must consider the circumstances, the legitimacy of restrictions, the available alternatives, and the effect on the reliability of the research.

Failure to provide a requested file does not automatically establish data fabrication or require retraction.

22. Licensing of Data, Software, and Research Materials

The Creative Commons Attribution 4.0 International License governing NOIS scholarly articles does not automatically establish the licensing status of separately distributed datasets, research software, or third-party materials.

Authors must identify the applicable rights and licenses for research objects they share.

For original research datasets, CC0 or an appropriate open data license may be used where legally and ethically permissible.

For software, an appropriate recognized open-source license is recommended when redistribution and reuse are authorized.

Restricted or proprietary materials must be accompanied by clear access and usage conditions.

Authors must not claim to license data, software, or other material for which they do not control the relevant rights.

Third-party datasets and software must be used and cited according to their applicable terms.

Licensing information must be sufficient for readers to determine whether they can legally access, reproduce, modify, or redistribute the material.

23. Data and Software Citation

Datasets, software packages, computational models, protocols, and other research objects that materially support a manuscript must be acknowledged and cited appropriately.

Citations should identify the creators, title or resource name, version, year, repository or distributor, and persistent identifier where available.

Authors must distinguish between citing an article that describes a dataset and citing the specific dataset actually analyzed.

When software versions materially affect the results, the exact version used should be identified.

Where research objects have DOIs, the DOI should be included in the relevant citation or availability statement.

The journal supports accurate citation of data and software as scholarly contributions deserving appropriate credit.

References to inaccessible or unspecified resources must not be presented as sufficient documentation of reproducibility.

24. Supplementary Materials and Publication Files

Supplementary materials must be relevant, properly labeled, and consistent with the main manuscript.

They may include extended proofs, source code, executable examples, benchmark results, data dictionaries, simulation configurations, additional tables, and other supporting research objects.

Essential methodological information needed to understand the principal findings must not be concealed in inaccessible or undocumented supplementary materials.

Files should use appropriate formats and include sufficient explanations for their contents to be interpreted independently.

The journal encourages accessible and widely supported formats where practicable.

Large or complex research objects should normally be placed in appropriate repositories rather than provided solely as manuscript attachments.

Authors must verify that supplementary files contain no unnecessary personal identifiers, confidential credentials, proprietary information disclosed without authority, or material that would compromise the journal's double-anonymous review process.

Submission files intended only for editorial assessment must be distinguished from material approved for public release.

Supplementary files accepted for publication must be included or linked accurately in the final publication record.

25. Double-Anonymous Review and Confidential Research Materials

NOIS uses double-anonymous peer review for manuscripts selected for external scientific evaluation.

Data and code repositories may reveal author identities through repository ownership, commit histories, account names, file metadata, or acknowledgments.

Authors must identify these risks before submitting resources for reviewer assessment.

When relevant, the editorial office may arrange access to anonymized research packages, confidential reviewer links, controlled-access repositories, or other appropriate materials.

Author-identifying records and confidential declarations must not be disclosed to reviewers when they are intended solely for authorized editorial use.

Scientific verification requirements do not justify unnecessary disclosure of research participants' personal information or confidential institutional records.

Reviewers must treat unpublished datasets, software, experimental results, and research documentation as confidential.

Reviewers must not redistribute, publish, or exploit confidential research materials obtained through the peer-review process.

26. Editorial Assessment of Reproducibility

The journal evaluates reproducibility requirements according to the nature and scientific significance of the submitted work.

During initial screening, editors should check the presence and adequacy of relevant availability statements and identify obvious omissions in methodological reporting.

During scientific peer review, reviewers assess whether the evidence, arguments, and available supporting materials are sufficient to evaluate the claims.

For theoretical work, this may involve examining assumptions, proof structure, definitions, and mathematical consistency.

For computational research, it may involve examining code, algorithms, parameter settings, dataset sources, or benchmark comparisons.

For artificial intelligence research, it may involve assessing model documentation, evaluation design, data leakage risks, and relevant performance evidence.

Where justified, the journal may request additional source materials or arrange independent methodological, statistical, mathematical, or computational verification.

Such verification must be documented accurately.

The journal does not claim that every published article has undergone independent code execution, complete experimental reproduction, or formal proof verification unless the relevant procedures have actually been performed.

Scientific peer review and successful independent reproduction are distinct forms of assessment.

27. Responsibilities of Authors, Reviewers, and Editors

Authors

Authors are responsible for the accuracy, provenance, documentation, accessibility statements, lawful use, and scientific integrity of the data, code, and materials supporting their claims.

They must respond reasonably to legitimate requests for clarification or verification and must not misrepresent the availability of research objects.

Reviewers

Reviewers should evaluate whether sufficient information has been provided to assess the relevant mathematical, methodological, computational, and empirical claims.

Reviewers must identify substantive concerns and distinguish established errors from uncertainties requiring clarification.

They must respect confidentiality, avoid unauthorized reuse of submitted materials, and disclose relevant conflicts of interest.

Editors

Editors are responsible for applying reproducibility requirements fairly and proportionately, selecting appropriate expert assessment where necessary, and ensuring that publication decisions reflect scientific evidence and ethical standards.

Editors must not require irrelevant data or unnecessary analyses solely to delay a decision or obtain access to a researcher's unpublished work.

28. Reproducibility Verification and Independent Assessment

When independent verification is warranted, the journal may adopt a proportionate assessment process.

The process may include:

Availability Check: Confirming that the cited research objects can be accessed under the stated conditions.

Documentation Check: Assessing whether the materials contain sufficient instructions, version information, and descriptions.

Mathematical Verification: Examining selected theoretical claims, derivations, or proof steps.

Computational Execution: Running available code or a representative test case when resources and expertise permit.

Result Comparison: Comparing reproduced outputs with reported values under appropriate tolerances and documented conditions.

Methodological Review: Obtaining specialist advice when the validity of methods or statistical analysis is uncertain.

Restriction Assessment: Evaluating whether justified access limitations permit an adequate alternative verification procedure.

Any claim that a study has passed independent reproduction or computational verification must be supported by an actual documented assessment.

The journal must not issue verification claims or badges solely because authors uploaded files or supplied an availability statement.

29. Non-Compliance and Research Integrity Concerns

Incomplete documentation, inaccessible materials, inconsistent results, or concerns about data or code integrity require proportionate editorial assessment.

Before publication, appropriate actions may include requesting clarification, correction, additional documentation, a reproducible example, revised analysis, or further independent review.

Where essential findings cannot be evaluated reliably, rejection may be justified on scientific grounds.

After publication, material errors or reliability concerns may require a correction, expression of concern, retraction, or another appropriate action under the journal's post-publication policies.

The journal distinguishes incomplete reporting, legitimate resource restrictions, honest computational errors, and deliberate misconduct.

An unavailable repository link or missing source file is not automatically proof of fabrication.

However, knowingly false data availability statements, manipulated research objects, fabricated outputs, or deliberate obstruction of legitimate verification may warrant a formal integrity investigation.

Affected authors must normally be given an appropriate opportunity to respond.

Any decision must consider the evidence and its implications for the published findings.

30. Post-Publication Availability and Continuing Responsibility

The responsibility to maintain accurate research availability information continues after publication.

Authors should notify the editorial office when they discover that a material repository link, archived dataset, or supporting code release is incorrect or has become unavailable.

Where feasible, authors should arrange appropriate restoration or replacement of access without compromising the integrity of the original research record.

Material changes to research objects must be documented using suitable versioning procedures.

The journal may update availability information through an appropriate correction when the change affects the ability to interpret or verify published findings.

The DOI and publication history of the original article must remain accurate.

Formal corrections, expressions of concern, and retractions are handled in accordance with the journal's published procedures.

31. Persistent Metadata and Long-Term Preservation

NOIS supports accurate identification and continued discoverability of data, software, and other scholarly research objects associated with its publications.

Where relevant persistent identifiers are available, they should be included in article metadata and reference information.

The publisher may register appropriate data and software citations or relationships through Crossref using supported metadata mechanisms.

The use of a repository, DOI, or versioned release does not itself guarantee permanent preservation of the underlying materials.

Authors should select suitable repositories and maintain accurate links and descriptive information.

The journal's Portico preservation arrangements concern the content covered by its actual preservation agreement and deposits.

Such arrangements must not be represented as automatic preservation of every externally hosted dataset, software repository, or research model referenced by an article.

32. Policy Review and Implementation

The journal and publisher are responsible for implementing this policy through clear author instructions, appropriate editorial procedures, and accurate public information.

Requirements will be reviewed when relevant disciplinary standards, research methods, legal obligations, or professional guidance change.

Material changes must be communicated transparently and reflected consistently in the journal's submission requirements and editorial policies.

The journal encourages the progressive improvement of reproducible research practices while recognizing that verification procedures must remain proportionate to the nature of each scholarly contribution.

Publication under this policy signifies that the article has undergone the journal's applicable editorial assessment. It does not constitute a guarantee that every computational result has been independently reproduced or that every mathematical claim has been formally verified beyond the documented evaluation.

Related Policies and Official Resources

Journal: Neutrosophic Optimization and Intelligent Systems (NOIS)

Publisher: Sciences Force LLC, United States

Editorial Office: nois@sciencesforce.com

Research Integrity Reports: ethics_reporting@sciencesforce.com

Publisher Data, Code, Materials and Reproducibility Policy: https://sciencesforce.com/publisher

Information for Authors: https://ojs.sciencesforce.com/index.php/nois/information_for_authors

Editorial Policies: https://ojs.sciencesforce.com/index.php/nois/policies

Publication Ethics Statement: https://ojs.sciencesforce.com/index.php/nois/publication_ethics

Official Submission Page: https://ojs.sciencesforce.com/index.php/nois/about/submissions

COPE — Publication Ethics Guidance: https://publicationethics.org/   

Center for Open Science — TOP Guidelines: https://www.cos.io/initiatives/top-guidelines

FAIR Guiding Principles: https://www.go-fair.org/fair-principles/

Crossref — Data and Software Citation: https://www.crossref.org/documentation/schema-library/markup-guide-metadata-segments/data-citation  

Official Journal Website: https://ojs.sciencesforce.com/index.php/nois/index