Glossary of research publishing terms

An abstract is a short, self-contained summary of a research work that states the question, the method, the main findings and the conclusion. Search engines and academic indexes use the abstract to decide what a record is about, so a record without a real abstract is hard to find.

Also called: author accepted manuscript (AAM), postprint

An accepted manuscript is the author's final version of a paper after peer review and acceptance, before the publisher's copy-editing, typesetting and branding. Many publishers allow authors to share the accepted manuscript in a repository, sometimes after an embargo.

Read more: Open access explained

An article processing charge is a fee some open-access journals charge to publish an article, usually paid by the author, their institution or their funder. Diamond open-access journals charge no APC.

Read more: Gold open access

BibTeX is a plain-text format and tool for managing references in LaTeX documents. Each entry has a type, such as @article or @misc, a citation key and fields such as author, title and year. The newer biblatex package adds entry types such as @dataset and @software.

Read more: Citing datasets and software

Bronze open access describes an article that is free to read on the publisher's website but carries no open license, so the publisher can withdraw free access at any time and reuse rights are not granted.

Read more: Open access explained

Also called: Creative Commons Attribution

CC BY is the Creative Commons Attribution license. It lets anyone copy, share, adapt and build on a work, including commercially, provided they credit the creator, link to the license and indicate whether changes were made. CC BY 4.0 is the current version.

Read more: Choosing an open license

Also called: Creative Commons Zero, public domain dedication

CC0 is a Creative Commons tool with which a rights holder waives all copyright and related rights in a work, as far as the law allows, placing it as close to the public domain as possible. CC0 is widely used for datasets, because attribution is then handled by scholarly citation rather than by a legal condition.

Read more: Licenses for data

The Citation File Format is a plain-text YAML file, conventionally named CITATION.cff and placed in a software repository's root, that tells people how to cite the software. GitHub reads the file and shows a "Cite this repository" option.

Read more: Citing software

Also called: Highwire Press tags

Citation metadata tags are HTML meta elements, such as citation_title, citation_author and citation_publication_date, that describe a scholarly document in machine-readable form. Google Scholar, Zotero, Mendeley and other tools read them to import correct bibliographic details. The citation_* names originate with HighWire Press.

Read more: Getting indexed in Google Scholar

A citation style is a set of rules for formatting references, such as APA, MLA, Chicago, Harvard, Vancouver or IEEE. The underlying facts (authors, year, title, publisher, identifier) are the same in every style; only order and punctuation change.

Creative Commons licenses are standard public copyright licenses that let creators grant permissions in advance. They combine four elements: Attribution (BY), ShareAlike (SA), NonCommercial (NC) and NoDerivatives (ND). Creative Commons licenses cannot be revoked for copies already obtained under them.

Read more: Comparing the licenses

Crossref is a not-for-profit DOI registration agency for scholarly publishing. Publishers register DOIs and metadata for journal articles, books, conference papers, preprints and other content with Crossref, which makes that metadata openly available.

Read more: What is a DOI?

A data availability statement is a short section in a paper that tells readers where the data underlying the findings can be found and under what conditions, ideally with a persistent identifier and a formal citation of the dataset.

DataCite is a not-for-profit DOI registration agency focused on research data, software and other research outputs. Data repositories such as Zenodo register DOIs through DataCite.

Read more: What is a DOI?

A dataset is a structured collection of data produced or used in research, such as measurements, survey responses, code outputs or images, published so others can verify, reuse or extend the work. A dataset can be cited in its own right, separately from any paper that analyses it.

Read more: How to cite a dataset

Also called: platinum open access

Diamond open access is publishing in which a journal or platform makes articles free to read and charges authors nothing to publish. Costs are covered by institutions, societies, funders or volunteers.

Read more: Open access explained

A DOI is a persistent identifier for a digital object, written as 10.prefix/suffix and resolved through https://doi.org/. The registrant keeps the DOI pointing at the current location of the object, so a citation using the DOI keeps working when a website changes. The DOI system is standardised as ISO 26324.

Read more: What is a DOI, and do you need one?

An embargo is a period set by a publisher after publication during which an author may not make a version of the article, usually the accepted manuscript, openly available in a repository.

Read more: Green open access

The FAIR principles state that research data and metadata should be Findable, Accessible, Interoperable and Reusable. They were published in Scientific Data in 2016 and are widely referenced by funders and data repositories.

Gold open access means a journal publishes an article openly on its own platform at the time of publication, usually under a Creative Commons license. Many gold open-access journals charge an article processing charge.

Read more: Open access explained

Google Scholar is Google's search engine for scholarly literature. It indexes papers it can crawl from publishers, repositories and university sites, groups versions of the same work, and counts citations between them.

Read more: Getting indexed in Google Scholar

Also called: self-archiving

Green open access means the author deposits a version of their work, often the accepted manuscript, in an open repository, so it is free to read regardless of how the journal publishes it.

Read more: Open access explained

A hybrid journal is a subscription journal that makes individual articles open access when an article processing charge is paid, while the rest of the journal stays behind a paywall.

Read more: Hybrid and bronze open access

An institutional repository is an archive run by a university or research organisation to collect, preserve and openly share the research outputs of its members. Subject repositories serve a discipline instead, and generalist repositories accept any field.

Read more: Repository

An ISSN (International Standard Serial Number) is an eight-digit identifier for a serial publication such as a journal, written as two groups of four digits, for example 1234-5679. Print and online editions of the same journal have different ISSNs.

Metadata is structured information that describes a research output: its title, authors, date, abstract, keywords, license, identifiers and files. Search engines, aggregators and reference managers work from metadata, so its accuracy decides whether a work can be found and cited correctly.

The Model Context Protocol is an open protocol, introduced by Anthropic in November 2024, through which AI applications connect to external tools and data sources. Life of Research runs a read-only MCP server so AI assistants can search and cite its records.

Read more: Developer documentation

OAI-PMH (Open Archives Initiative Protocol for Metadata Harvesting) is a standard protocol, at version 2.0, through which repositories expose their metadata so that aggregators and discovery services can collect it in bulk.

Read more: Developer documentation

Open access means research outputs are free to read online without a subscription or payment, usually with a license that also permits reuse. The main routes are green (repository), gold (journal) and diamond (journal without author fees).

Read more: Open access explained

An open license is a standard license that lets anyone use, share and usually adapt a work without asking permission, under stated conditions such as attribution. Creative Commons licenses for text and data, and licenses such as MIT or Apache 2.0 for software, are open licenses.

Read more: Choosing an open license

An ORCID iD is a free, persistent identifier for an individual researcher, written as a 16-character number such as 0000-0002-1825-0097 and displayed as a link to orcid.org. It distinguishes researchers with similar names and links them to their work across publishers, funders and repositories.

Read more: ORCID explained

Peer review is the evaluation of a manuscript by independent experts before a journal accepts it. Common models are single-anonymous, double-anonymous and open review. Peer review checks a paper's methods and claims but does not guarantee they are correct.

Read more: Preprints vs journal articles

A persistent identifier is a long-lasting reference to a digital object, person or organisation that keeps resolving even when locations change. DOIs identify works, ORCID iDs identify researchers and ROR IDs identify research organisations.

Plan S is an initiative of cOAlition S, a group of research funders, launched in 2018. It requires research funded by its members to be published open access immediately, under an open license, preferably CC BY.

Read more: Open access explained

A preprint is a complete research manuscript shared publicly before, or in parallel with, formal peer review. Preprints are citable and usually carry a DOI or other identifier, but they have not been certified by a journal.

Read more: Preprints vs journal articles

A preprint server is a repository dedicated to preprints, usually for one discipline, such as arXiv for physics, mathematics and computer science or bioRxiv for biology. Most preprint servers screen submissions but do not peer review them.

A reference manager is software that stores bibliographic records and formats citations and bibliographies, such as Zotero, Mendeley or EndNote. Browser connectors read a page's citation metadata to save a record with one click.

A research repository is a service that stores research outputs, gives each one a stable address and metadata, and makes it publicly available. Life of Research is a generalist open-access repository: it accepts papers, preprints, datasets, theses and software from any discipline.

Read more: Depositing step by step

A retraction is a formal notice by a journal or publisher that a published article should no longer be relied upon, typically because of serious error or misconduct. The article usually stays online, marked as retracted, with the notice explaining why.

Rights retention is a strategy in which authors keep the right to share their accepted manuscript openly, typically by stating at submission that a CC BY license applies to it, so a later publishing agreement cannot remove that right.

Read more: Before you choose a license

RIS is a plain-text format for exchanging bibliographic records between reference managers. Each line holds a two-letter tag and a value, such as TY for the item type, AU for an author and DO for a DOI.

Self-archiving is an author depositing their own work, typically a preprint or accepted manuscript, in a repository to make it openly available. It is the mechanism behind green open access.

Read more: Green open access

Signposting is a convention for adding typed links, such as cite-as, author, license and describedby, to HTTP Link headers of scholarly landing pages, so machines can find a work's persistent identifier, metadata and files without parsing HTML.

The version of record is the final, published form of an article as made available by the publisher, after copy-editing and typesetting. It is the version that should be cited when the citation relies on the published text.

Read more: Article versions