RESEARCH MISCONDUCT

Jurnal Sistem Cerdas

1. General Policy on Research Misconduct

Jurnal Sistem Cerdas is committed to maintaining the highest standards of research integrity, academic honesty, transparency, accountability, and ethical scholarly publishing.

The journal recognizes that the reliability and credibility of scientific publications depend on the responsible conduct of research and the accurate reporting of scientific findings.

Research misconduct undermines scientific knowledge, damages public trust, and compromises the integrity of the scholarly publication process.

Jurnal Sistem Cerdas adopts a zero-tolerance approach to deliberate research misconduct while recognizing the importance of distinguishing intentional wrongdoing from honest errors, legitimate scientific disagreements, and methodological limitations.

This policy establishes the principles, procedures, and editorial measures for identifying, investigating, and addressing suspected research misconduct before and after publication.

The policy is informed by internationally recognized publication ethics principles, including guidance from the Committee on Publication Ethics (COPE).

2. Definition of Research Misconduct

Research misconduct refers to serious violations of accepted scientific and ethical standards during the proposal, conduct, evaluation, reporting, submission, or publication of research.

The journal recognizes three fundamental forms of research misconduct:

Fabrication: Making up research data, experimental results, observations, or scientific evidence and presenting them as genuine findings.

Falsification: Manipulating research materials, equipment, procedures, datasets, images, or results, or omitting information in a manner that causes the research record to be materially misleading.

Plagiarism: Using another individual's ideas, methods, words, findings, or intellectual contributions without appropriate acknowledgment or attribution.

Other violations of publication ethics, including inappropriate authorship, duplicate publication, and citation manipulation, are also addressed under this policy.

Not every ethical violation constitutes research misconduct in the narrow institutional or legal sense. The journal evaluates concerns according to their nature, seriousness, supporting evidence, and applicable standards.

3. Types of Research Misconduct

3.1. Data Fabrication

Data fabrication involves creating fictitious research data, observations, experimental records, or results that were not actually obtained through legitimate research activities.

Examples include:

Inventing experimental datasets or research observations.
Reporting machine learning performance without conducting the stated experiments.
Creating fictitious survey responses or research participants.
Reporting fabricated accuracy, precision, recall, or F1-score values.
Presenting simulated findings as actual experimental results without disclosure.
Creating nonexistent baseline experiments or comparative results.

Authors must ensure that all reported research findings are supported by genuine and appropriately documented evidence.

3.2. Data Falsification

Data falsification involves manipulating research procedures, materials, datasets, experimental configurations, or results in ways that misrepresent the actual findings.

Examples include:

Altering experimental outcomes to support a preferred hypothesis.
Modifying research data without appropriate scientific justification.
Intentionally omitting unfavorable results to mislead readers.
Manipulating evaluation metrics or confusion matrices.
Misrepresenting training, validation, or testing procedures.
Concealing significant experimental limitations.
Misleadingly altering figures, graphs, or scientific images.

Legitimate data cleaning, preprocessing, transformation, augmentation, or statistical analysis does not constitute falsification when scientifically justified, accurately documented, and transparently reported.

3.3. Plagiarism

Plagiarism occurs when authors present the intellectual contributions of others as their own without proper acknowledgment.

Examples include:

Copying text from published or unpublished sources without appropriate citation.
Reproducing research ideas, algorithms, or methodologies without acknowledgment.
Using figures, tables, or research results without proper attribution.
Paraphrasing existing publications without citing the original sources.
Submitting manuscripts substantially derived from another person's work without acknowledgment.
Presenting another researcher's software or computational framework as an original contribution.

All submitted manuscripts may undergo similarity screening using appropriate plagiarism detection software.

The editorial team evaluates similarity reports contextually, considering citation practices, source attribution, disciplinary conventions, and the nature of textual overlap.

A similarity percentage alone is not conclusive evidence of plagiarism.

3.4. Duplicate and Redundant Publication

Duplicate publication occurs when substantially the same research is published more than once without appropriate disclosure, justification, or acknowledgment.

Redundant publication involves presenting substantially overlapping research findings as separate original publications without sufficient scientific distinction.

Examples include:

Submitting substantially identical manuscripts to multiple journals simultaneously.
Republishing previously published research without appropriate disclosure.
Dividing research findings into multiple minimally differentiated manuscripts to increase publication counts.
Presenting previously published experimental findings as new results.
Concealing substantial overlap with earlier publications.

Authors must disclose relevant previous publications, conference papers, and preprints.

Extensions of previously published research may be considered when they contain meaningful new scientific contributions and appropriately acknowledge the earlier work.

3.5. Authorship Misconduct

Authorship misconduct involves misrepresenting the contributions or responsibilities of individuals involved in a research project.

Prohibited practices include:

Honorary or Gift Authorship: Listing individuals as authors without sufficient scholarly contribution.

Ghost Authorship: Omitting contributors who satisfy the journal's authorship criteria.

Coercive Authorship: Pressuring researchers to include individuals who do not qualify for authorship.

Purchased Authorship: Buying, selling, or exchanging authorship positions.

Unauthorized Authorship Changes: Adding, removing, or rearranging authors without appropriate agreement and editorial approval.

All authors must comply with the journal's Authorship Criteria and approve the final manuscript.

3.6. Citation Manipulation

Citation manipulation involves intentionally influencing citation practices to artificially increase citation counts or other scholarly metrics.

Examples include:

Including irrelevant references solely to increase citation counts.
Excessive and unjustified self-citation.
Coercive citation requests by editors or reviewers.
Coordinated citation exchanges among researchers or journals.
Adding references unrelated to the manuscript's scientific argument.

References must be relevant, accurate, verifiable, and scientifically justified.

3.7. Image and Figure Manipulation

Scientific images, figures, charts, and visual representations must accurately reflect the underlying research findings.

Prohibited practices include:

Manipulating images to conceal or fabricate evidence.
Duplicating or reusing images while misrepresenting their origins.
Altering graphical results to exaggerate performance improvements.
Removing relevant visual information without disclosure.
Presenting misleading charts or fabricated experimental visualizations.

Appropriate image enhancement and visualization procedures are permitted when scientifically justified and transparently reported.

4. Misconduct in Artificial Intelligence and Machine Learning Research

Jurnal Sistem Cerdas recognizes that artificial intelligence and machine learning research presents specific risks related to data integrity, experimental reproducibility, model evaluation, and computational transparency.

Authors must ensure that AI-related research is conducted and reported responsibly.

4.1. Dataset Manipulation

Research misconduct may involve intentional misrepresentation of dataset composition, quality, sources, or experimental use.

Examples include:

Falsely reporting the number of training or testing samples.
Concealing intentional data leakage between training and test datasets.
Misrepresenting synthetic datasets as real-world observations.
Falsifying dataset annotations or class distributions.
Misrepresenting dataset ownership or access authorization.
Intentionally concealing known dataset limitations to mislead readers.

4.2. Experimental Result Manipulation

Authors must accurately report model performance and experimental procedures.

Prohibited practices include:

Fabricating model accuracy or other performance metrics.
Reporting experiments that were never conducted.
Claiming nonexistent comparisons with baseline models.
Altering experimental results to demonstrate artificial improvements.
Concealing deliberate manipulation of test conditions.
Misrepresenting model training procedures or computational settings.

Selective model development, hyperparameter tuning, and experimental optimization are legitimate research practices when accurately reported and evaluated using appropriate methods.

4.3. Misrepresentation of Algorithmic Novelty

Authors must accurately describe the originality and scientific contribution of proposed algorithms, models, and computational frameworks.

Research misconduct may involve falsely claiming ownership of existing algorithms, deliberately concealing substantial reuse of previous methods, or presenting another researcher's implementation as an independently developed innovation.

The use of established algorithms is acceptable when appropriately acknowledged and accompanied by a meaningful scientific contribution.

4.4. Reproducibility and Experimental Transparency

Authors should provide sufficient details to support independent scientific evaluation.

Relevant information may include:

Dataset characteristics and acquisition procedures.
Data preprocessing methods.
Computational model architecture.
Training and evaluation configurations.
Hyperparameters and optimization procedures.
Baseline comparison methods.
Performance evaluation metrics.
Software libraries and computational environments.
Random seeds and experimental repetition, where applicable.

Failure to achieve reproducibility does not automatically constitute misconduct. However, deliberate misrepresentation or concealment of material experimental information may raise research integrity concerns.

5. Misuse of Generative Artificial Intelligence

Jurnal Sistem Cerdas supports responsible use of generative artificial intelligence tools while requiring human accountability for scientific content.

AI systems cannot be designated as authors because they cannot assume the ethical and scholarly responsibilities required of human researchers.

Authors must disclose substantive use of generative AI tools in accordance with the journal's Publication Ethics policy.

Prohibited AI-related practices include:

Generating fabricated research findings and presenting them as genuine.
Creating fictitious citations or bibliographic references.
Producing misleading scientific images or experimental evidence.
Using AI-generated text to conceal plagiarism.
Misrepresenting AI-generated methodologies or results as independently validated research.
Submitting scientific claims without appropriate verification.
Using AI tools in ways that violate research confidentiality, privacy, or intellectual property requirements.

The responsible use of AI tools for language assistance, programming support, or research activities is not inherently misconduct.

Human authors remain fully responsible for verifying the accuracy, originality, reliability, and ethical compliance of their work.

6. Ethical Violations in Human and Sensitive Data Research

Research involving human participants, personal data, biometric information, healthcare records, educational records, or sensitive datasets must comply with applicable ethical and legal requirements.

Potential violations include:

Collecting personal data without required authorization or consent.
Conducting research without required ethical approval.
Disclosing confidential participant information.
Misrepresenting the status of ethical approval.
Using restricted datasets without appropriate permission.
Deliberately exposing identifiable information without a lawful and ethical basis.

The journal may request evidence of ethical approval, informed consent, or data use authorization where applicable.

7. Responsibilities of Authors

Authors are responsible for ensuring the integrity, accuracy, originality, and transparency of their research.

Authors must:

Conduct research according to accepted scientific and ethical standards.
Report research methods and findings accurately.
Maintain appropriate research records and supporting documentation.
Provide genuine and verifiable citations.
Disclose conflicts of interest and funding sources.
Obtain required ethical approvals and permissions.
Ensure appropriate authorship attribution.
Disclose relevant use of generative AI tools.
Cooperate with reasonable editorial inquiries concerning research integrity.
Promptly notify the editorial office of significant errors or misconduct concerns discovered after submission or publication.

All authors share responsibility for ensuring that questions regarding research integrity are appropriately investigated and resolved.

8. Responsibilities of Editors

Editors are responsible for maintaining the integrity of the editorial process and addressing suspected research misconduct fairly and objectively.

Editors must:

Evaluate allegations without prejudice.
Protect confidential information.
Assess the credibility and seriousness of concerns.
Request clarification or supporting documentation when appropriate.
Avoid conflicts of interest during misconduct assessment.
Consult relevant experts or institutions when necessary.
Maintain appropriate records of editorial investigations.
Determine editorial actions based on available evidence.
Implement corrections, expressions of concern, or retractions when justified.
Protect the integrity of the scholarly publication record.

Editors must not presume guilt solely because an allegation has been submitted.

9. Responsibilities of Reviewers

Reviewers are expected to assist the editorial team in identifying potential research integrity concerns.

Reviewers should notify editors when they identify suspected plagiarism, fabricated data, manipulated results, inappropriate citations, or other ethical concerns.

Reviewers must provide objective and evidence-based assessments.

Reviewers must not independently publicize allegations or disclose confidential manuscript information.

Potential misconduct concerns must be communicated through the journal's authorized editorial channels.

10. Reporting Suspected Research Misconduct

Jurnal Sistem Cerdas accepts credible reports of suspected research misconduct from authors, reviewers, editors, readers, institutions, and other stakeholders.

Reports should be submitted to the editorial office through the journal's official communication channels.

A report should contain, where available:

The title or identification number of the affected manuscript or article.
A clear description of the suspected misconduct.
Supporting documents or evidence.
Identification of the relevant research findings or publication content.
Contact information for follow-up communication, when the reporting individual is willing to provide it.

Anonymous reports may be considered when sufficient information and credible evidence are provided.

The editorial team will assess reports fairly and protect confidentiality to the extent reasonably possible.

11. Procedures for Investigating Research Misconduct

Jurnal Sistem Cerdas applies a structured approach to evaluating allegations of research misconduct.

Stage 1: Receipt and Documentation

The editorial office receives and records the allegation.

Relevant information, supporting evidence, and communication records are documented for further assessment.

Stage 2: Preliminary Assessment

The Editor-in-Chief or an authorized editor evaluates whether the allegation is sufficiently specific, credible, and relevant to warrant further investigation.

The preliminary assessment may involve reviewing the manuscript, publication history, similarity reports, research documentation, or other available evidence.

Unsubstantiated allegations may be closed without further action.

Stage 3: Author Notification and Clarification

When appropriate, the corresponding author and other relevant parties are contacted and given an opportunity to respond.

Authors may be asked to provide original data, experimental records, explanations, ethical approval documents, or other supporting materials.

The editorial team must consider the authors' responses objectively.

Stage 4: Further Investigation or Institutional Referral

If concerns remain unresolved, the editorial team may seek additional expert assessment or refer the matter to the relevant institution, research organization, or responsible authority.

The journal generally does not replace institutions or competent authorities in conducting formal investigations into individual researchers.

The journal remains responsible for determining appropriate editorial action concerning submitted or published articles.

Stage 5: Editorial Evaluation

The editorial team evaluates the available evidence, the authors' explanations, and any relevant institutional findings.

The seriousness of the concern, its impact on the scientific record, and the reliability of the manuscript's findings are considered.

Stage 6: Editorial Action

Appropriate action is determined based on the available evidence and applicable publication ethics standards.

Possible actions include requesting corrections, rejecting a manuscript, publishing a correction, issuing an expression of concern, or retracting an article.

Stage 7: Documentation and Follow-Up

The editorial office records the outcome and maintains relevant documentation.

Where necessary, the journal communicates with authors, institutions, indexing services, or other appropriate parties.

Investigations should be conducted without unnecessary delay, although the time required may vary according to the complexity of the case.

12. Actions Before Publication

When research misconduct is suspected during manuscript submission, editorial screening, or peer review, the journal may take appropriate measures.

Possible actions include:

Requesting clarification from the authors.
Requesting original research data or supporting documentation.
Suspending the editorial evaluation.
Conducting additional similarity or integrity screening.
Consulting independent experts.
Referring serious concerns to the relevant institution.
Rejecting the manuscript when justified.

Rejection for scientific weakness or methodological inadequacy does not necessarily imply research misconduct.

13. Actions After Publication

If serious concerns arise after an article has been published, the editorial team will evaluate the matter in accordance with the journal's Publication Ethics policy.

Depending on the circumstances, the journal may issue:

Correction

A correction is appropriate when significant errors are identified but the overall scientific findings remain reliable.

Expression of Concern

An expression of concern may be published when serious questions regarding an article's reliability or integrity remain unresolved.

Retraction

Retraction may be appropriate when findings are unreliable because of fabrication, falsification, plagiarism, serious errors, unethical research, or other substantial integrity concerns.

Editorial Notice

An editorial notice may be issued when additional clarification is necessary to inform readers about the status of the published research.

The journal aims to maintain a transparent and accurate scholarly record.

14. Retraction Policy

A published article may be retracted when reliable evidence demonstrates that its findings are fundamentally unreliable or that serious publication ethics violations have occurred.

Retraction decisions must consider the nature of the problem, the available evidence, and the implications for the scientific record.

Retraction notices should:

Clearly identify the affected article.
Explain the reason for retraction.
Identify the party issuing the retraction.
Remain publicly accessible.
Be appropriately linked to the original publication.
Clearly indicate the retracted status of the article.

Retracted articles should remain identifiable in the journal's archive, with their retracted status clearly indicated.

Retraction is intended to correct the scientific literature and protect readers rather than serve primarily as a punitive measure.

15. Institutional Notification and Cooperation

The journal may communicate with authors' institutions or relevant research organizations when serious allegations require investigation beyond the editorial team's authority.

Institutional notification may be appropriate when:

Serious data fabrication or falsification is suspected.
Authorship disputes cannot be resolved through editorial procedures.
Ethical approval or participant protection concerns arise.
Systematic publication misconduct is suspected.
Supporting research records require institutional verification.

The journal may consider institutional findings when determining editorial action.

However, editorial decisions concerning manuscript acceptance, correction, or retraction remain the responsibility of the journal.

16. Confidentiality and Fair Treatment

All allegations of research misconduct must be handled with appropriate confidentiality, fairness, impartiality, and respect for the rights of involved parties.

The journal will not automatically treat unverified allegations as established misconduct.

Authors must be given a reasonable opportunity to respond to substantive concerns.

Editors and reviewers involved in a misconduct assessment must disclose relevant conflicts of interest.

The journal will seek to protect individuals who raise concerns in good faith from inappropriate retaliation within the journal's editorial processes.

Deliberately false or malicious allegations may also be subject to appropriate editorial consideration.

17. Appeals Against Misconduct Decisions

Authors who disagree with an editorial decision related to alleged research misconduct may submit a formal appeal.

The appeal should include:

The manuscript or article identification number.
The decision being appealed.
A clear explanation of the grounds for appeal.
Relevant supporting evidence.
Any additional documentation necessary for reassessment.

Appeals should be evaluated by an editor or qualified individual who is sufficiently independent of the original decision.

The journal may uphold, modify, or reconsider the original decision when justified by the evidence.

An appeal does not automatically suspend a necessary action to protect the integrity of the scholarly record.

18. Relationship Between Research Misconduct and Publication Fees

Jurnal Sistem Cerdas applies an Article Processing Charge (APC) of IDR 500,000 per accepted article.

Payment of the APC does not exempt authors from research integrity requirements or guarantee publication.

Editorial decisions concerning suspected misconduct must remain independent of financial considerations.

The journal may reject a manuscript or issue a post-publication correction or retraction when justified, regardless of whether publication fees have been paid.

Any financial consequences, including refunds, are governed by the journal's applicable Publication Fees and Refund Policy.

19. Prevention of Research Misconduct

Jurnal Sistem Cerdas encourages preventive measures to promote responsible research practices and reduce the risk of scientific misconduct.

These measures include:

Clear publication ethics policies.
Transparent authorship and contribution requirements.
Appropriate similarity and originality screening.
Rigorous and independent peer review.
Transparent reporting of research methodologies.
Appropriate management of conflicts of interest.
Responsible use of artificial intelligence tools.
Data management and reproducibility practices.
Clear procedures for reporting ethical concerns.
Continuous improvement of editorial quality control.

The journal encourages researchers to adopt responsible scientific practices throughout the research lifecycle.

20. Commitment to Research Integrity

Jurnal Sistem Cerdas is committed to maintaining a trustworthy and ethical scholarly publishing environment.

The journal expects authors, editors, reviewers, and publishing personnel to uphold the principles of honesty, transparency, objectivity, accountability, and scientific responsibility.

Research misconduct allegations will be evaluated fairly and consistently, with editorial action proportionate to the available evidence and the seriousness of the concern.

Through these commitments, Jurnal Sistem Cerdas aims to protect the reliability of published research and support the responsible advancement of intelligent systems and computational science.

21. References and Ethical Guidelines

This policy is informed by internationally recognized research integrity and publication ethics resources, including:

Committee on Publication Ethics (COPE)

https://publicationethics.org/

COPE Core Practices

https://publicationethics.org/core-practices

COPE Retraction Guidelines

https://publicationethics.org/guidance/guideline/retraction-guidelines

International Committee of Medical Journal Editors (ICMJE)

https://www.icmje.org/recommendations/

Office of Research Integrity (ORI)

https://ori.hhs.gov/

Authors, editors, reviewers, and other stakeholders are expected to comply with the journal's official Publication Ethics, Authorship Criteria, Peer Review Process, and Editorial Process policies.