Reviewer Guidelines
The KORA Journal of Applied AI & Multidisciplinary Studies (KJAI) relies on the expertise and commitment of qualified reviewers to maintain the quality, integrity and academic standard of published research. A new journal is only as good as the people who read for it.
You will usually be reviewing half a paper
Cross-domain manuscripts go to one methods reviewer and one domain reviewer. When you are invited, you are told which half you are being asked to assess. Say clearly in your report where your competence ends — that is what lets the editor read the two reports together and see what neither of you covered. It is a mark of a good reviewer, not a failing.Role of the Reviewer
Reviewers are responsible for:
- assessing the originality and significance of the research
- evaluating the soundness of the method, the experimental design and the analysis
- judging whether the evidence presented actually supports the claims made
- checking that baselines, splits, metrics and reported uncertainty are appropriate
- assessing whether the method is described well enough to be reimplemented
- checking the domain grounding, where a domain is involved, and whether a specialist would trust the result
- checking that code and data are available, or that their unavailability is explained
- assessing clarity, structure and presentation
- identifying the strengths and weaknesses of the manuscript
- providing constructive suggestions for improvement
- recommending a publication decision to the editor
Conducting the Review
When reviewing a manuscript, consider:
- relevance to the journal's aims and scope
- novelty and contribution to knowledge — including a negative result that genuinely closes something off
- methodological validity and the appropriateness of the evaluation
- whether the comparison is fair: were baselines tuned with a comparable budget?
- whether uncertainty is reported and what it means
- whether the data is described sufficiently — version, size, licence, provenance, preprocessing
- accuracy of results and interpretation, and whether the abstract overstates them
- quality of writing and organisation, and whether both audiences can follow it
- adequacy and accuracy of references
- ethical compliance — consent, ethics approval, disclosure of generative AI use, and discussion of foreseeable harms
What not to hold against a manuscript
A result that is modest, a negative result, a study on a small or unfashionable dataset, work from a low-resource setting, and an author who is plainly early in their career are none of them reasons to reject. Nor is "this could have used a larger model" a review comment unless you can say why the conclusion depends on it. Judge whether the claims made are supported by the evidence given.Reviewing code and data
Where a manuscript releases code or data, you are welcome — not required — to look at it, and the invitation will say whether the authors have released anything. If you do look, treat it exactly as you treat the manuscript: confidential, not to be run against your own data, not to be reused, and not to be kept after the review. Report what you found in your report, not to the authors directly.
Review Recommendations
• AcceptPublishable as it stands
• Minor RevisionAccepted in principle; the editor checks the revision
• Major RevisionResubmission invited and re-reviewed
• RejectNot publishable here, with the specific reason
Reviewers advise; editors decide. Your recommendation is weighed alongside the second report and the editor's own reading, and you will be sent the other report and the final decision so you can see how a colleague from a different field read the same manuscript.
Timeliness
Reviewers are asked to complete a report within 21 days, with a reminder at day 14. If you cannot meet the deadline, tell the editorial office promptly — an extension is almost always possible, and a manuscript kept waiting in silence is the single most common failure in scholarly publishing. A fast decline is more useful to us than a slow acceptance, and declining never affects future invitations.
Confidentiality
All manuscripts, code, data and review materials are confidential documents. Reviewers must not share, discuss or use their content for personal or professional advantage, and must not upload any of it to a generative AI service or hosted model.
Conflict of Interest
Decline any assignment where a conflict of interest exists — personal, professional, financial or institutional. In a field where a great deal of research is funded or employed by a small number of companies, a commercial conflict is as real as an academic one: declare employment, equity, consultancy and funding. If you recognise the author part-way through a review and a conflict emerges, tell the editor at once rather than finishing the report.
Constructive Feedback
Reviews should be respectful, professional and focused on scholarly improvement. Personal criticism of an author is not acceptable. Write the report you would want to receive: specific, evidenced, pointing to a location in the text, and clear about what would fix each problem.
Review Ethics
Peer review only works if everyone involved behaves as though it matters. These are the obligations a KJAI reviewer accepts on agreeing to review, and they are also printed on the report form.
Confidentiality
- A manuscript under review is an unpublished, confidential document. Do not share it, forward it, or discuss its content with anyone outside the editorial process.
- Do not use its findings, data, code or ideas in your own work before publication.
- Do not upload a manuscript, its data or its code — in whole or in part — to any generative AI service, hosted model, or third-party tool that retains what is sent to it. This is a breach of confidentiality, not a grey area, and it includes pasting an abstract into a chatbot to help you summarise it.
- You may use an AI tool to polish the language of your own report, but the judgements must be yours, and the use must be declared to the editor. A report that is visibly generated rather than written is not a review.
- Do not run, redistribute or build on code or data supplied with a manuscript for any purpose other than assessing that manuscript, and delete your copy when the review is complete.
- Do not contact the author directly, even if you work out who they are.
Objectivity and Conflicts
- Judge the work, not the author's institution, nationality, career stage, employer or presumed identity.
- Decline where you have collaborated with, supervised or been supervised by a likely author within three years, work in the same department or company, have a close personal relationship, or stand to gain from the outcome.
- Decline where the manuscript competes directly with your own unpublished work.
- Declare a commercial interest — employment, equity, consultancy or funding — in a company whose product or model the manuscript evaluates. It does not disqualify you; it tells the editor where to be careful.
- Declare any connection to The Korean Academy. It does not disqualify you — it tells the editors which manuscripts not to send you.
Competence
Accept only what you are competent to assess. Reviewing outside your expertise helps nobody, and it is the fastest way for a small journal to lose credibility. Because this journal publishes cross-domain work, partial competence is normal and expected: if you can assess the method but not the clinical setting, or the economics but not the model, accept and tell the editor exactly which part you are qualified to judge, so the second reader can cover the rest.
Reporting a Concern
If you suspect plagiarism, duplicate publication, fabricated or manipulated data, missing consent, an undisclosed conflict or undisclosed AI use, describe what you saw and where you saw it — do not accuse, and do not investigate on your own. The editor investigates, following the journal's publication ethics policy, and the author is given the evidence and a chance to respond.
Anonymity and Recognition
- Review is double-blind. Your identity is never disclosed to authors, and the manuscript you receive carries no author details.
- Reviewers are named, with permission, in the annual acknowledgement published each December. You may decline to be named and it makes no difference to whether we invite you.
- A certificate of review is available on request for your own records or ORCID profile.
Become a Reviewer
KJAI is building its reviewer pool across all of the journal's subject areas — artificial intelligence methods, the application domains the journal publishes in, and responsible and human-centred AI. Applications are open and are reviewed as they arrive. Everyone who wishes to review for the journal — including people invited personally by an editor — completes the same application form, so that the basis on which reviewers are selected is a matter of record.
Two requirements
Requirement 1
An ORCID iD
Every reviewer must have an ORCID iD. It is free, takes about two minutes to create at orcid.org, and it is how your review record stays attached to you rather than to an email address you may one day lose. Applications without one cannot be processed.
Requirement 2
Any two of eight criteria
You must meet at least two of the eight criteria below, and give evidence for each one you claim. They are deliberately varied: research degrees, publications, industry practice, open-source work and recognised domain expertise all count, and no single route is privileged over another.
Eligibility criteria — meet any two
- A doctorate, completed or in progress, in computer science, artificial intelligence, statistics, engineering, or in a domain field in which you apply AI.
- A master's degree in one of the journal's subject areas.
- A bachelor's degree in a relevant field, together with demonstrable further specialisation in the area you wish to review.
- At least one peer-reviewed publication — journal article, book chapter or full conference paper — within the journal's scope.
- At least two years' professional experience building, deploying or evaluating AI or data systems in industry, clinical practice, government or a research group.
- Prior peer-review experience for a journal, conference or academic press.
- A substantive maintained open-source contribution, released dataset or released model within the journal's scope.
- Recognised domain expertise — clinical, legal, educational, engineering, policy or equivalent — in a field where AI is being applied, including professional certification or licensure where relevant.
A note to doctoral students, early-career researchers and practitioners
If you are doing or have recently finished a research degree in AI or in a field where AI is applied — or you build and deploy these systems for a living without an academic post — we strongly encourage you to apply. You do not need a long publication record, and you do not need a doctorate. Criteria F2, F3, F5, F7 and F8 exist precisely so that capable people without a conventional CV can qualify.
Practitioners are worth saying something about specifically. A great deal of what goes wrong in applied AI papers is invisible from inside a laboratory — the split that leaks, the label that is recorded after the decision it is meant to predict, the metric that nobody in the field actually uses. If you have spent two years putting models into production in a hospital, a bank, a factory or a school, you can see those things, and this journal needs you on its reviewer list.
Early-career reviewers are matched only to manuscripts within their stated competence, are never asked to assess methods they have not been trained in, and receive the other reviewer's report alongside the decision — which is, in practice, one of the better ways to learn how the field judges work.
What we ask in return is honesty about what you can and cannot assess. Section G of the form asks you to tick your areas: please be conservative. Declining a manuscript that falls outside your competence is a mark of a good reviewer, not a failing.
The application form
The form below is the complete application. Copy it into a document or into the body of an email, fill in every section, and send it with your CV. Do not delete sections — write “N/A” where something does not apply.
KJAI Reviewer Application Form
Ten sections: contact details, identifiers and profiles, education, current position, technical background, the eligibility criteria, areas of expertise, publications and outputs, availability, and the reviewer declarations.
FORM
kjai-reviewer-application-form.txt 279 lines · press Copy above, then fill it in
================================================================================
KORA JOURNAL OF APPLIED AI & MULTIDISCIPLINARY STUDIES (KJAI)
코라 응용 인공지능 및 다학제 연구 저널
REVIEWER APPLICATION FORM
Published by The Korean Academy
================================================================================
HOW TO SUBMIT
1. Complete every section below. Do not delete any section; write "N/A" where
something does not apply.
2. Attach your CV (PDF preferred).
3. Email both to: admission@thekoreanacad.com
4. Subject line: KJAI Reviewer Application - [Full Name] - [Country]
BEFORE YOU START -- two things are required of every applicant
* An ORCID iD. Registration is free and takes two minutes at orcid.org.
Applications without an ORCID iD cannot be processed.
* At least TWO of the eight eligibility criteria in Section F.
Date of application: ____________________
--------------------------------------------------------------------------------
SECTION A -- PERSONAL AND CONTACT DETAILS
--------------------------------------------------------------------------------
A1. Title (Mr / Ms / Mx / Dr / Prof / other):
A2. Full name (as it should appear in the reviewer acknowledgement):
A3. Preferred name form for correspondence:
A4. Email address:
A5. Alternative email:
A6. Phone (with country code):
A7. Country of residence:
A8. City:
A9. Time zone:
--------------------------------------------------------------------------------
SECTION B -- IDENTIFIERS (B1 is mandatory)
--------------------------------------------------------------------------------
B1. ORCID iD (required): https://orcid.org/____-____-____-____
B2. Google Scholar profile URL:
B3. DBLP / Semantic Scholar profile URL:
B4. GitHub, GitLab or other code profile URL:
B5. Personal or departmental web page:
--------------------------------------------------------------------------------
SECTION C -- EDUCATION
--------------------------------------------------------------------------------
List your highest qualification first. Add rows as needed. Include qualifications
currently in progress, marking them "in progress" with the expected year.
C1. Degree / Level :
Field of study :
Institution :
Country :
Year awarded / expected :
Thesis or dissertation title (if any):
C2. Degree / Level :
Field of study :
Institution :
Country :
Year awarded / expected :
Thesis or dissertation title (if any):
C3. Degree / Level :
Field of study :
Institution :
Country :
Year awarded / expected :
--------------------------------------------------------------------------------
SECTION D -- CURRENT POSITION
--------------------------------------------------------------------------------
D1. Current role / job title:
D2. Department or team:
D3. Institution, company or organisation:
D4. Country:
D5. Since (month / year):
D6. Sector (university / research institute / industry / government / clinical /
NGO / independent):
D7. If you are a student, state your programme, year of study, and the name of
your supervisor (we do not contact them without asking you first):
--------------------------------------------------------------------------------
SECTION E -- TECHNICAL BACKGROUND
--------------------------------------------------------------------------------
E1. Which best describes you? (tick all that apply)
[ ] Primarily a methods researcher (machine learning, AI, statistics)
[ ] Primarily a domain researcher who applies AI in that domain
[ ] Both, roughly equally
[ ] A practitioner who builds and deploys AI systems in production
[ ] A researcher on the ethical, legal, social or policy side of AI
E2. Application domains you work in (medicine, education, engineering, climate,
finance, law, linguistics, social science, other):
E3. Can you assess a paper's statistics and experimental design? Yes / No
E4. Can you read and evaluate released code? Yes / No
E5. Programming languages and frameworks you are fluent in:
E6. Can you review a manuscript written in English? Yes / No
E7. Any other language you can review in:
--------------------------------------------------------------------------------
SECTION F -- ELIGIBILITY (tick AT LEAST TWO, and give the evidence)
--------------------------------------------------------------------------------
[ ] F1. A doctorate (completed or in progress) in computer science, artificial
intelligence, statistics, engineering, or in a domain field in which you
apply AI.
Evidence:
[ ] F2. A master's degree in one of the journal's subject areas.
Evidence:
[ ] F3. A bachelor's degree in a relevant field, together with demonstrable
further specialisation in the area you wish to review.
Evidence:
[ ] F4. At least one peer-reviewed publication -- journal article, book chapter
or full conference paper -- within the journal's scope.
Evidence (full citation):
[ ] F5. At least two years' professional experience building, deploying or
evaluating AI or data systems in industry, clinical practice, government
or a research group.
Evidence:
[ ] F6. Prior peer-review experience for a journal, conference or academic press.
Evidence (name of venue; you need not name the manuscripts):
[ ] F7. A substantive maintained open-source contribution, released dataset or
released model within the journal's scope.
Evidence (repository or archive link, and your role):
[ ] F8. Recognised domain expertise -- clinical, legal, educational, engineering,
policy or equivalent -- in a field where AI is being applied, including
professional certification or licensure where relevant.
Evidence:
Number of criteria ticked: ______ (minimum 2)
--------------------------------------------------------------------------------
SECTION G -- AREAS OF EXPERTISE
--------------------------------------------------------------------------------
G1. Tick every subject area you are competent to review. Please be honest and
conservative: reviewing outside your competence helps nobody.
METHODS
[ ] Machine learning, deep learning and representation learning
[ ] Statistical learning theory and optimisation
[ ] Natural language processing and large language models
[ ] Computer vision and image analysis
[ ] Speech, audio, signal and time-series modelling
[ ] Reinforcement learning, planning and decision-making
[ ] Knowledge representation, graphs and neuro-symbolic methods
[ ] Generative models, retrieval-augmented and agentic systems
[ ] Efficient, edge, federated and privacy-preserving learning
[ ] Data engineering, MLOps and production deployment
[ ] Evaluation, benchmarking and uncertainty quantification
APPLICATION DOMAINS
[ ] Health, medicine and the life sciences
[ ] Education and the learning sciences
[ ] Engineering, manufacturing and robotics
[ ] Environment, climate, agriculture and energy
[ ] Finance, economics, business and operations
[ ] Law, governance and public policy
[ ] Language, translation, digital humanities and cultural heritage
[ ] Computational social science
[ ] Security, forensics and critical infrastructure
RESPONSIBLE AND HUMAN-CENTRED AI
[ ] Fairness, bias and auditing
[ ] Explainability and interpretability
[ ] Privacy, data protection and provenance
[ ] Robustness, safety and failure analysis
[ ] Human-AI interaction and cognitive science
[ ] AI ethics, regulation and standards
[ ] Other (specify): ______________________________________
G2. Five keywords describing what you can review:
G3. Methods you are competent to assess (for example: experimental design,
statistical inference, causal inference, benchmarking methodology,
qualitative and mixed methods, human-subject study design, systematic
review method, formal analysis):
G4. Would you be willing to review the code or data released with a manuscript,
where the authors have released it? Yes / No
--------------------------------------------------------------------------------
SECTION H -- SELECTED PUBLICATIONS OR OUTPUTS (up to five, most relevant first)
--------------------------------------------------------------------------------
Full citations. Released datasets, models and maintained software count as
outputs -- give the persistent identifier or repository. Write "None yet" if you
have none; this does not disqualify you where other criteria are met.
H1.
H2.
H3.
H4.
H5.
--------------------------------------------------------------------------------
SECTION I -- REVIEW EXPERIENCE AND AVAILABILITY
--------------------------------------------------------------------------------
I1. Have you reviewed for a journal, conference or press before? Yes / No
I2. If yes, where, and roughly how many manuscripts:
I3. How many manuscripts can you review per year? (we ask for no more than 3)
I4. Any period in the coming year when you will be unavailable:
I5. Would you be willing to review a manuscript at short notice if a reviewer
withdraws? Yes / No
--------------------------------------------------------------------------------
SECTION J -- DECLARATIONS (initial each line)
--------------------------------------------------------------------------------
____ The information given in this form is accurate and complete.
____ I will keep every manuscript sent to me confidential; I will not share it,
discuss it with anyone outside the editorial process, or use its content
before publication.
____ I will not upload a manuscript, its data, or its code -- in whole or in
part -- to any generative AI service, hosted model or third-party tool
that retains what is sent to it. I understand this is a breach of
confidentiality, not a grey area.
____ I will not run, redistribute or build on code or data supplied with a
manuscript for any purpose other than assessing that manuscript.
____ I will declare any conflict of interest and will decline to review the
work of a close colleague, co-author within three years, classmate,
student, supervisor or anyone in my own department or company.
____ I will decline promptly, rather than accept and delay, when a manuscript
falls outside my competence or I cannot meet the deadline.
____ I will write reviews that are professional and constructive, and will not
make personal criticism of an author.
J1. Do you consent to being named in the journal's annual reviewer
acknowledgement? (You may say no; it makes no difference to whether we
invite you.) Yes / No
J2. Do you have any connection to The Korean Academy -- as a student, alumnus,
employee, teacher or contractor? This does not disqualify you; it tells us
which manuscripts we must not send you.
Yes / No. If yes, describe:
J3. Do you have a commercial interest -- employment, equity, consultancy or
funding -- in a company whose products this journal publishes work on?
This does not disqualify you either; it tells us where to be careful.
Yes / No. If yes, describe:
--------------------------------------------------------------------------------
SIGNATURE
--------------------------------------------------------------------------------
Signed: ____________________________ Date: ____________________
(A typed name is accepted for an emailed application.)
--------------------------------------------------------------------------------
ATTACHMENTS CHECKLIST
--------------------------------------------------------------------------------
[ ] This completed form
[ ] Curriculum vitae (PDF preferred) -- required
[ ] One publication, preprint, writing sample or repository link (optional, but
genuinely helpful if you have no publication record yet)
================================================================================
Send to: admission@thekoreanacad.com
Subject: KJAI Reviewer Application - [Full Name] - [Country]
You will receive an acknowledgement within 3 to 5 working days. Successful
applicants are added to the reviewer database and matched to manuscripts in
their stated areas. We aim not to invite any reviewer more than three times
a year.
================================================================================
How to submit
Send your application to
admission@thekoreanacad.com
Subject line:
KJAI Reviewer Application - [Full Name] - [Country]- Complete the form in full. Every section, with evidence for each eligibility criterion you claim.
- Attach your CV — PDF preferred. This is required, not optional. It should show your education, any research or professional posts, and your publications and released artefacts if you have them.
- Attach supporting evidence if you have it — a publication, a preprint, a writing sample or a repository link. Optional, but genuinely helpful if you have no publication record yet.
- Email everything to admission@thekoreanacad.com using the subject line above.
What happens next
| Stage | What happens | Timeline |
|---|---|---|
| Acknowledgement | We confirm receipt of your application and CV | 3–5 working days |
| Assessment | An editor checks the ORCID, the two criteria and the evidence given | 2–3 weeks |
| Outcome | You are added to the reviewer database, or told what is missing | Within 3 weeks |
| First invitation | Sent when a manuscript matches your stated areas — which may be some months away | As manuscripts arrive |
If your application is not accepted
We will tell you why, and what would change the answer. Most unsuccessful applications fail on one of two things: no ORCID iD, or only one criterion met with no evidence attached. Both are fixable, and you are welcome to apply again as soon as they are.
Being listed as a KJAI reviewer requires your consent, and no one is named publicly without it. You may withdraw from the reviewer pool at any time by writing to the editorial office.