Author Guidelines

These guidelines apply to every manuscript submitted to QAIRU Journal, published by Qazaq AI Research University. Read them before you prepare the manuscript. A submission that does not meet them is returned at the editorial check, before review.

Before you submit

Check the scope. QAIRU Journal publishes work in four areas: artificial intelligence and machine learning; robotics and intelligent systems; applied AI; and computational methods and high-performance computing. Read Aims and Scope and name the area your manuscript addresses in the cover letter. A manuscript outside these areas is returned without review.

Choose the article type. The journal publishes Research Articles, Review Articles, Short Communications and Editorials. Each has its own word limit and structure; see the table below. If you are unsure which type fits your work, ask the editors before you submit.

Confirm the work is original. The manuscript must not have been published before and must not be under consideration by another journal. If any part of the material has appeared elsewhere in any form, describe it in the cover letter and give the identifier or the link.

Confirm authorship. Every author must meet the criteria set out in Authorship and Contributions, every author must have approved the version you submit, and no one who qualifies as an author may be omitted. Agree the order of authors before submission, not after review.

Confirm the language. The journal works in English only, at every stage. The manuscript must be in publication-quality English. Language editing is the author's responsibility, and the editors may return a manuscript for language editing before review.

Article types

Word limits exclude the abstract, references, tables and figure captions.

If your manuscript must exceed a limit, request permission in the cover letter before you submit and explain why the length is necessary. A manuscript that exceeds a limit without prior agreement is returned.

Manuscript structure

Research Articles follow IMRAD. Each part must contain the following.

• Title. Specific and informative: state what was studied and how. Avoid abbreviations and avoid declaring novelty in the title.
• Abstract. See "Abstract and keywords" below.
• Keywords. See "Abstract and keywords" below.
• Introduction. The problem and why it matters; what prior work has established and where it stops; and your contribution, stated explicitly. End the section with the contribution in plain terms, as a short list where there is more than one.
• Methods. Enough detail for an independent researcher to reproduce the work: data sources and their licences, preprocessing steps, model architectures, hyperparameters and how they were selected, random seeds, training and inference hardware, software versions, and the evaluation protocol including metrics, baselines and statistical treatment.
• Results. The findings, without interpretation. Report measures of variability, such as standard deviations or confidence intervals over repeated runs, wherever a comparative claim is made. A single-run number is not sufficient evidence of an improvement.
• Discussion. Interpretation of the results, comparison with prior work, and an explicit statement of the limitations of the study in its own paragraph. Do not restate the Results.
• Conclusion. What the work establishes and what follows from it. No new results.
• Declarations. Placed before the references, in the order given below.
• References. IEEE style.
• Appendices. Material that supports the argument but would interrupt it. Label A, B, C and cite each one in the text.

Review Articles replace Methods with a section that states the search strategy — the sources searched, the query strings used, and the date of the last search — together with the criteria used to include and exclude sources and the number of sources at each stage. Short Communications follow IMRAD but may merge Results and Discussion. Editorials have no prescribed structure.

Files to submit

1. Anonymised main manuscript, in DOCX or PDF, containing everything from the title to the appendices with all identifying information removed.
2. Title page, as a separate file, holding all identifying information (see below).
3. Figures, as separate files where possible, one file per figure, named by figure number.
4. Tables, in editable form, either in the main manuscript or as a separate editable file. Never as images.
5. Supplementary material, one file per item, each with a short description, and each cited in the main text.
6. Cover letter, stating the article type, the subject area, why the work suits the journal, any prior appearance of the material, and any request to exceed a word limit.

LaTeX submissions are sent as a compiled PDF for review; the source files (.tex, .bib, figure files, class and style files) are supplied on acceptance.

Preparing an anonymised manuscript

Review is double-blind: reviewers do not learn who wrote the manuscript, and authors do not learn who reviewed it. See Peer Review Process. The main file must therefore carry nothing that identifies you.

• Remove all author names, affiliations, e-mail addresses and ORCID identifiers from the main file.
• Move acknowledgements, funding details with grant numbers, and ethics approval details to the title page, and keep them there only.
• Refer to your own prior work in the third person: "Author and colleagues showed", not "we showed in our earlier work". Keep the citations themselves — a reviewer needs them.
• Remove identifying information from the file properties: author, organisation, last-modified-by, comments and tracked changes.
• Replace links to personal pages, institutional repositories, code repositories and project sites with an anonymised link or with the note that the link will be supplied on acceptance.
• Do not name your institution, laboratory, computing cluster or internal dataset where the name would identify you. Describe it generically and add the detail on acceptance.

Title page

Submit the title page as a separate file containing the following.

Abstract and keywords

The abstract is at most 250 words, unstructured — one paragraph, no headings — with no citations and no undefined abbreviations. It must state the problem, the approach taken, the main result including its principal quantitative outcome, and why that result matters. Write it so that it stands alone, because for many readers it is the only part they will read.

Give 4 to 6 keywords. Choose the terms a reader searching this field would actually enter. Avoid words that already appear in the title where you can, avoid abbreviations unless they are standard in the field, and separate keywords with semicolons.

Formatting

Figures and tables

• Number figures and tables in separate sequences, each in order of first mention, and cite every one in the text.
• Give each a self-contained caption: the figure or table must be understandable without the body text. Figure captions go below the figure, table captions above the table.
• Supply raster images at 300 dpi or more at the size at which they will be printed. Vector formats are preferred for plots and diagrams.
• Check readability at the printed size, which is about 8 cm for a single column and about 17 cm across the full width. No text inside a figure may be smaller than 8 point.
• Use colour so that the figure still reads in greyscale and for readers with colour vision deficiency: distinguish series by shape, line style or direct labelling as well as by hue.
• Supply tables in editable form, never as images. Table footnotes go below the table, marked with superscript letters.
• A reused figure or table requires written permission from the rights holder that is compatible with CC BY 4.0. Supply the permission at submission and credit the source in the caption. See Licensing and Copyright.
• Do not manipulate image data. Adjustments to brightness, contrast or colour balance must be applied to the whole image and disclosed in the caption. Where fields or lanes from different sources are grouped, mark the join with a dividing line.

References

Use IEEE style.

• Cite in the text with numbers in square brackets, in order of first appearance: [1], [2], [4]–[6].
• Order the reference list by citation number, not alphabetically.
• Every reference must be cited in the text, and every citation must appear in the list.
• Include the DOI wherever one exists.
• Cite preprints as preprints, with the repository, the identifier and the version.
• Cite datasets and software in the reference list, with their persistent identifiers and version numbers.
• Verify every reference against the primary source. Do not cite a work you have not read, and never carry over a reference produced by a generative AI tool without confirming that it exists and says what you claim. See Generative AI Policy.

Worked examples, with placeholder authors:

Journal article [1] A. B. Author, C. D. Author, and E. F. Author, "Title of the paper," Abbrev. Title of Journal, vol. 12, no. 3, pp. 45-58, 2026, doi: 10.0000/example.1234.

Conference paper [2] A. B. Author and G. H. Author, "Title of the conference paper," in Proc. Name of Conference (ABBR), City, Country, 2026, pp. 101-110, doi: 10.0000/example.5678.

Dataset [3] A. B. Author and C. D. Author, "Title of the dataset," Name of Repository, version 1.2, 2026. [Dataset]. doi: 10.0000/example.9012.

Required declarations

Place these before the reference list, each under its own heading, in the order below. Write "Not applicable" only where the item genuinely does not apply.

Author Contributions. State what each author did, using CRediT roles and author initials. See Authorship and Contributions. Model wording: "Conceptualisation, A.B.A. and C.D.A.; methodology, A.B.A.; software, C.D.A.; validation, E.F.A.; formal analysis, A.B.A.; investigation, C.D.A.; data curation, C.D.A.; writing — original draft, A.B.A.; writing — review and editing, all authors; visualisation, C.D.A.; supervision, E.F.A.; funding acquisition, E.F.A. All authors have read and agreed to the submitted version of the manuscript."

Conflicts of Interest. Declare, for every author, each financial and non-financial interest that a reader could reasonably regard as capable of influencing the work. See Conflicts of Interest. Model wording: "The authors declare no conflicts of interest." Or: "A.B.A. is employed by [organisation], which funded the study; the funder had no role in the design of the study, in the analysis, or in the decision to publish."

Funding. Name every funder in full and give the grant numbers. Model wording: "This work was supported by [funder] under grant number [number]." Or: "This research received no external funding."

Declaration of Generative AI Use. State whether generative AI tools were used in preparing the manuscript or in the research itself, which tools and versions, and for what purpose. An AI tool cannot be an author, and the authors remain responsible for everything in the manuscript. See Generative AI Policy. Model wording: "During the preparation of this manuscript the authors used [tool, version] to [purpose]. The authors reviewed and edited the output and take full responsibility for the content of the publication." Or: "The authors declare that no generative AI tools were used in the preparation of this manuscript."

Data Availability Statement. Say where the data and the code behind the findings can be obtained, on what conditions and under which licence. See Data Availability. Model wording: "The data that support the findings of this study are openly available in [repository] at [DOI], under [licence]." Or: "The data are not publicly available because [reason]; they are available from the corresponding author on reasonable request."

Ethics approval and informed consent. Required where the research involves people, their identifiable data, or animals. Model wording: "The study was approved by [name of committee], approval number [number], on [date]. Written informed consent was obtained from all participants." Or: "Not applicable."

Acknowledgements. Thank the people and facilities that contributed without meeting the authorship criteria. Obtain the permission of everyone you name.

How to submit

Submit through the online submission system. Manuscripts are not accepted by e-mail.

What happens next:

1. Editorial check. The editors verify scope, article type, the completeness of files and declarations, formatting and anonymisation, and screen the manuscript with a licensed similarity-detection service. A manuscript is returned at this stage for a missing declaration, a mismatch with the scope, a formatting failure, identifying information left in the main file, or a similarity finding.
2. Peer review. At least two independent reviewers, double-blind. Reviewers answer an invitation within one week and return a report within three weeks.
3. First decision. Four to six weeks from submission. The handling editor takes the decision; the Editor-in-Chief is responsible for the final decision.
4. Revision. 30 days for a minor revision, 60 days for a major revision. Return the revised manuscript with a point-by-point response to every comment and the changes marked in the text.

The full procedure is set out in Peer Review Process.

After acceptance

Supply the LaTeX source files if you submitted a compiled PDF. Proofs are sent to the corresponding author for approval and must be returned within 72 hours. Only errors are corrected at proof stage: typographical mistakes and errors introduced in production. New results, new references and rewritten passages are not accepted at proof.

The article is published under CC BY 4.0 with copyright retained by the authors, and a DOI is assigned via Crossref upon publication. No charges apply at any stage — see Fees.

Templates

Manuscript template (Word, .docx) and Manuscript template (LaTeX, .zip)

Using a template is recommended but not mandatory, provided the manuscript meets the requirements set out above.

Submission checklist

• [ ] The manuscript falls within the scope of the journal, and the subject area is named in the cover letter.
• [ ] The article type is chosen, and the manuscript is within its word limit — or permission to exceed it is requested in the cover letter.
• [ ] The manuscript is in publication-quality English.
• [ ] The work is original, is not published elsewhere, and is not under consideration by another journal. A preprint or a conference abstract does not count as prior publication, but it must be disclosed in the cover letter.
• [ ] Every author meets the authorship criteria, has approved this version, and the order of authors is agreed.
• [ ] The main file is anonymised: no names, affiliations, e-mails, ORCID identifiers, acknowledgements, funding details or ethics details, and no identifying information in the file properties.
• [ ] Self-citations are written in the third person and are retained, not deleted.
• [ ] A separate title page carries the full title, running title, all authors, affiliations with ROR identifiers where available, an e-mail and an ORCID for every author, the marked corresponding author, acknowledgements, funding and ethics details.
• [ ] The abstract is at most 250 words, unstructured, with no citations and no undefined abbreviations.
• [ ] There are 4 to 6 keywords.
• [ ] Formatting matches the table above: A4, single column, 1.5 spacing, 11 or 12 point serif, numbered pages, continuous line numbers, SI units, numbered equations.
• [ ] Figures and tables are numbered, cited in the text, captioned so that they stand alone, supplied at 300 dpi or in vector form, and readable in greyscale; tables are editable, not images.
• [ ] Written permission compatible with CC BY 4.0 is held for every reused figure or table.
• [ ] References follow IEEE style, are numbered in order of first appearance, include DOIs where they exist, and every one has been verified against the primary source.
• [ ] All required declarations are present before the reference list, in the prescribed order.
• [ ] The Declaration of Generative AI Use names the tools and their purpose, or states that none were used.
• [ ] All files are ready: main manuscript, title page, figures, tables, supplementary material and cover letter.

---

Effective date: 2026-09-01 · Version: 1.0 · Questions: journal@qairu.edu.kz