Publication Ethics

Commitment, Scope & Editorial Principles

ARCITECH: Journal of Computer Science and Artificial Intelligence is committed to responsible scholarly communication, methodological transparency, impartial assessment, and the reliability of the published record. This policy applies to research submissions, reviews, accepted and published articles, supplementary materials, and post-publication concerns.

Authors, editors, reviewers, editorial staff, and the publisher share responsibility for upholding research integrity. Editorial decisions shall rest on academic quality, originality, relevance, methodological soundness, and ethical compliance—never on author identity, affiliation, nationality, political or religious views, personal relationships, or ability to pay publication charges.

Responsibilities of Authors

All listed authors are accountable for the integrity of their contributions and shall cooperate with reasonable editorial requests to clarify a submission.

• Submit original and accurate work that is not simultaneously under consideration by another journal; disclose closely related papers, conference versions, preprints, and overlapping datasets.
• State the research problem, contribution, limitations, methods, assumptions, experimental setting, statistical analysis, and relevant negative or inconclusive findings accurately.
• Maintain appropriate research records, permissions, dataset provenance, and computational evidence that support the reported findings.
• Identify all funding sources, institutional and commercial affiliations, and relevant competing interests, including a declaration of no conflicts where applicable.
• Credit prior scholarship, third-party data, software, code, images, and other material correctly; obtain authorization for restricted content.
• Disclose errors promptly at any stage and cooperate with appropriate revisions, clarifications, corrections, or other post-publication notices.

Authorship, Contributorship & Acknowledgments

Authorship is reserved for individuals who made a substantive intellectual contribution to the conception, design, acquisition, analysis, or interpretation of the work; participated in drafting or critically revising it; approved the final version; and accept accountability for the contribution and integrity of the work. A contributor whose role does not meet authorship criteria should instead be acknowledged with permission.

• The corresponding author must confirm that all named authors consented to submission and approved the final version and author order.
• Guest, gift, honorary, ghost, purchased, and fictitious authorship are unacceptable.
• Authors may be asked to provide an individual contribution statement using recognized contributor roles, such as CRediT.
• Requests to add, remove, or reorder authors after submission must include an explanation and written agreement of all affected authors; disputes that cannot be resolved may be referred to the relevant institution.
• Automated systems, large language models, and other AI tools cannot be credited as authors.

Originality, Prior Dissemination & Citation Ethics

Duplicate submission and redundant publication are prohibited when substantially the same findings are presented as new without disclosure, justification, and permission where applicable. A disclosed preprint or conference paper is not automatically disqualifying; the editor assesses the material relationship, added scholarly contribution, copyright restrictions, and any applicable prior-publication policy.

• Disclose preprints, theses, conference proceedings, technical reports, and previously published analyses that materially overlap the submission.
• Clearly identify the manuscript’s original contribution relative to earlier versions and cite all relevant sources.
• Avoid inappropriate self-plagiarism, data slicing intended to disguise redundant publication, citation cartels, excessive coercive citations, and fabricated or unverifiable references.
• References must support the corresponding assertions. Cite primary sources where reasonably possible; never invent DOIs, publication details, or quotations.

Plagiarism, Falsification & Publication Manipulation

Unacceptable practices include unattributed copying or paraphrasing; fabrication or selective falsification of data, experimental observations, code outputs, or statistics; deceptive image alterations; manipulated citations; forged reviewer identities; fraudulent peer-review reports; and submissions acquired or assembled through paper-mill operations.

ARCITECH publicly states that Turnitin is used for similarity checking and refers to a 20% similarity limit. This figure is an administrative screening reference, not a scientific determination of plagiarism. Editors must examine the source and context of overlapping text, including quotations, methods, references, and properly cited reuse; low similarity likewise does not establish originality.

Data, Software & Computational Reproducibility

Computational research should provide sufficient methodological detail for independent technical assessment and, where lawful and feasible, replication. Authors shall state material limitations when full data or code release is prevented by privacy, security, licensing, or legitimate contractual restrictions.

• Describe dataset origin, collection period, sampling strategy, inclusion criteria, transformations, labeling methodology, missing-data treatment, and licensing or access restrictions.
• Report software and library versions, key dependencies, algorithms, model configuration, hyperparameters, initialization, random seeds, and major computing environment characteristics when material to replication.
• Explain training, validation, and testing protocols, performance metrics, confidence intervals or uncertainty estimates as appropriate, and relevant statistical assumptions.
• Provide an accessible code, data, or materials link when available, or a clear data/software availability statement explaining restrictions and reasonable access conditions.
• Do not fabricate a repository URL or claim data/code are publicly available unless the materials can actually be accessed.

Integrity in AI, Machine Learning & Data Science

For machine learning and artificial intelligence research, the reported methodology must permit assessment of the validity of model comparisons and generalization claims.

• Prevent and disclose potential information leakage between training, validation, test, temporal, user-level, or site-level partitions; explain any balancing, augmentation, or feature selection before evaluation.
• Report appropriate baselines, dataset sizes and class distributions, evaluation metrics suitable for the task, and meaningful ablation or sensitivity analyses when relevant.
• Do not claim broad deployment readiness from narrow synthetic datasets or a single benchmark; disclose dataset shift, distributional limitations, uncertainty, and risks of bias where applicable.
• Clearly distinguish actual observations from synthetic or simulated data and label illustrative or AI-generated visualizations accurately.
• For image, audio, and sensor research, preserve original evidence where feasible and disclose transformations that materially affect interpretation.

Cybersecurity, Sensitive Data & Responsible Disclosure

Security research involving networks, websites, devices, software vulnerabilities, system logs, user records, or threat telemetry must be conducted with appropriate authorization and safeguards.

• Explain legal or institutional authority for active probing, penetration testing, exploitation experiments, collection of access logs, or security testing of systems not owned by the authors.
• Remove or protect passwords, tokens, keys, session identifiers, personal identifiers, unpublished vulnerability details, and other information that could expose individuals or systems to harm.
• Describe ethical handling of dual-use data and a proportionate responsible-disclosure process when the publication could expose exploitable weaknesses.
• Use de-identified, aggregated, synthetic, or access-controlled materials where open publication would breach privacy, security commitments, or applicable laws.

Responsible Use of AI-Assisted and Generative Tools

Authors must disclose substantive use of generative AI or automated tools in preparing text, code, data analysis, generated images, or study design, excluding routine spelling and language correction. Disclosures should identify the tool, purpose, and extent of use in an appropriate manuscript section. AI output must be independently checked for factual, computational, citation, originality, and intellectual-property errors.

• AI tools cannot be authors, nor can AI-generated text be cited as a reliable scholarly source in place of the underlying evidence.
• Reviewers and editors must not upload confidential manuscripts, peer-review reports, or private correspondence to external AI systems where confidentiality cannot be assured.
• Reviewers must not delegate substantive scientific assessment or report writing to generative AI; any permitted limited editorial assistance must preserve confidentiality and be disclosed.
• Editors remain responsible for all assessments and must independently verify automated integrity alerts rather than treating AI-detection scores as definitive proof.

Ethical Approval, Human Participants & Consent

Research involving human participants, identifiable personal information, interviews, surveys, behavioral monitoring, biometric information, or sensitive digital traces must comply with applicable laws and institutional ethical standards. Where prior ethics review is required, authors must report the responsible body, approval or exemption, and relevant identifier where available.

• Obtain informed consent when required, including for identifiable quotations, photographs, recordings, or other details that could reveal a participant.
• Explain privacy protections, data minimization, retention, and access controls appropriate to the sensitivity of the materials.
• Never assume that publicly visible online activity or accessible technical logs may automatically be published without ethical assessment.
• For studies where formal ethics approval is not required, explain the applicable exemption or institutional position when the nature of the research raises a material ethics question.

Duties of Editors & the Publisher

Editors are accountable for fair, timely, and reasoned decisions. They shall apply the published scope and review procedures, protect confidential submissions, and avoid using unpublished information for personal advantage. The publisher supports editorial independence and the integrity of the scholarly record, including when a concern involves a journal officer.

• Assess submissions without discrimination or inappropriate influence from APC income, sponsor relationships, institutional status, or other non-scientific considerations.
• Assign competent, independent reviewers and maintain records of reviewer selection, recommendations, conflicts of interest, and editorial outcomes within the publishing workflow.
• Recuse any editor who is an author, close collaborator, or materially conflicted party; assign an independent, qualified handling editor.
• Take allegations seriously whether received before or after publication; correct the public record when warranted.

Peer Review, Reviewer Conduct & Confidentiality

ARCITECH describes a double-blind peer-review process and review by two subject-matter reviewers for manuscripts that pass editorial screening. Scientific reviews must be conducted independently, confidentially, constructively, and without discriminatory or personal remarks.

• Reviewers accept assignments only when qualified, sufficiently independent, and able to meet the agreed schedule; otherwise they promptly decline.
• Reviewers disclose financial, institutional, personal, or intellectual conflicts and must not use manuscripts, data, or findings for their own benefit.
• Review reports should evaluate scientific soundness, originality, clarity, reproducibility, analytical validity, and appropriate acknowledgment of prior work.
• Suspected plagiarism, duplicated publication, falsified experiments, or image/data irregularities should be reported confidentially to the editor with substantiating information.
• Reviewers must not contact authors privately, share confidential manuscripts, or submit third-party reviews in place of their own assessment.

Competing Interests, Funding & Editorial Independence

Authors, reviewers, editors, and journal officials must declare financial and nonfinancial relationships that could reasonably be perceived to influence interpretation or decisions. These may include grants, employment, consultancy, patents, personal relationships, academic rivalry, institutional leadership, and contractual obligations.

Relevant funding and sponsor roles must be reported. A declared conflict does not automatically disqualify a paper; the editor determines appropriate management measures, including additional independent evaluation or recusal. Manuscripts from journal officers or reviewers receive no preferential scientific treatment.

Intellectual Property, Images & Research Materials

Authors retain responsibility for lawful use of texts, images, tables, code, datasets, and third-party material. Material not owned by the authors must be credited and accompanied by required permissions or lawful reuse justification. Presentation changes must not misrepresent underlying evidence.

The journal currently states that authors retain copyright and articles are licensed under Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0). Licensing and permission requirements must remain consistent with the journal’s separately published copyright and open-access policies.

• All image adjustments, synthesis, generative augmentation, composites, or illustrative renderings that affect scientific interpretation must be described.
• Scientific images must not be deceptively assembled, duplicated, relabeled, selectively altered, or presented as an original observation when they are simulated or generated.

Reporting and Assessing Allegations of Misconduct

Concerns regarding manuscripts, peer review, published content, or editorial conduct may be reported by authors, readers, reviewers, staff, or institutions. Good-faith concerns will be considered irrespective of whether the reporter is named. Anonymous reports may be assessed when they contain sufficient verifiable evidence.

The handling editor or an unconflicted designated editor will make a proportionate preliminary assessment, preserve relevant records, and offer implicated parties a reasonable opportunity to respond when appropriate. Complex questions of primary research misconduct may be referred to authors’ institutions or other competent authorities; the journal does not substitute for an institutional investigation. Confidentiality and fair treatment will be maintained as far as practicable.

Corrections, Expressions of Concern & Retractions

The journal will evaluate credible post-publication concerns according to the reliability of the findings and the seriousness of the issue. An editorial correction is appropriate for material errors that can be reliably corrected without invalidating the central conclusions. An Expression of Concern may be issued where substantial unresolved concerns merit public notice. Retraction may be warranted where findings are unreliable, misconduct or unethical practices materially undermine the work, or peer review has been compromised.

• Notices will clearly identify the affected article, explain the reason to the extent responsible and legally appropriate, and link to the published record.
• Retraction does not generally mean silently deleting the article; the scholarly record and notice should remain discoverable and clearly marked.
• Authors are invited to cooperate with the assessment, but an unresolved dispute over authors’ agreement does not in itself prevent an appropriate editorial notice.
• Substantiated post-publication discussion and scholarly criticism may be submitted to the editorial office for evaluation and response.

Complaints, Appeals & Special Editorial Cases

A complaint or an appeal should provide the manuscript or article identifier, a concise account of the concern, and relevant evidence. Mere disagreement with a reviewer’s scientific opinion does not automatically establish misconduct. Allegations of procedural error, unmanaged conflict, discrimination, ethical breach, or demonstrable factual mistake will be considered.

Where practicable, an appeal shall be examined by an editor or publisher representative not responsible for the original contested decision. Concerns regarding the Editor-in-Chief or the editorial team should be marked for confidential escalation to publisher-level oversight through the official journal address. Special issues, guest-edited content, and manuscripts authored by editors are subject to the same core ethical standards and independent assessment as ordinary research articles.

Editorial decisions and appeal outcomes will be communicated through official channels, with an explanation appropriate to confidentiality obligations.

Contact, Related Policies & Guiding Standards

Questions, integrity reports, complaints, and requests for clarification may be directed to the ARCITECH Editorial Office. Please include the manuscript ID or DOI when known; do not send unnecessary personal information or sensitive security evidence in an unprotected email.

Related journal pages: About / Peer Review  ·  Author Guidelines  ·  Publication Fee.

Guiding frameworks: DOAJ Guide to Applying; COPE–DOAJ–OASPA–WAME Principles of Transparency and Best Practice (v4); COPE Core Practices; ICMJE Guidance on AI.

Use of these sources reflects policy alignment, not certification, membership, or endorsement by the listed organizations.