What the Acronym Extractor does
This extractor finds every acronym and abbreviation in a document - NHS, COPD, APIs, R&D, COVID-19 - matches each one to its spelled-out form, and tells you which are defined at first use, which are used before they are defined, which are never defined and which are expanded two different ways. It then exports a ready-made list of abbreviations as CSV, Markdown or HTML.
Expansions are found with the Schwartz and Hearst algorithm, a published method that matches the letters of "COPD" against the words before or inside a bracket, such as "chronic obstructive pulmonary disease (COPD)". It runs in your browser; the document is not uploaded.
How to use it
- Paste the document or drop a text or Markdown file. The table updates as you type.
- Read the findings first: acronyms used before their definition, acronyms defined two ways, and acronyms never spelled out, each with the line number of the first use.
- Decide whether widely known abbreviations such as UK, EU, PDF or NASA need defining for your readers. By default they are not counted as undefined; untick the box if your style guide requires every abbreviation to be spelled out.
- Type any missing expansions straight into the table. They are used in the exports, so you can build the abbreviations list without editing the document first.
- Download the list as CSV, Markdown or an HTML definition list that uses the abbr element, or copy it into your document's front matter.
Reading the results
Defined means the spelled-out form appears at or before the first use, in either "long form (ABC)" or "ABC (long form)" order.
Defined after first use means readers meet the acronym before its meaning. Move the definition to the first use; the finding shows the exact wording to use.
Two expansions means the same letters are defined with different words, such as ONS for two different organisations. Differences only in capitals, hyphens or punctuation - "Service Level Agreement" and "service-level agreement" - are not counted as a conflict.
Worked example: a health report with nine abbreviations
Load the health report. The extractor finds 9 acronyms: NHS, COPD, NICE, PR, GP, WHO, EU, BMI and ONS. Three are defined correctly at first use: NICE (National Institute for Health and Care Excellence), PR (pulmonary rehabilitation) and WHO, defined the other way round as "WHO (World Health Organization)".
COPD appears in the first sentence and is only spelled out in the second, so it is flagged as defined late. ONS is defined as both "Office for National Statistics" and "Online Network Service" - a conflict a reader cannot resolve.
NHS, GPs and BMI are never spelled out, so 3 acronyms are counted as undefined. EU is never spelled out either, but it is on the well-known list and is shown separately rather than counted - untick the option if your readers need it defined.
Why define abbreviations at all
An abbreviation saves the writer time and costs the reader some. Readers who do not know it must guess or search, and screen reader users may hear it spelled out letter by letter or read as a word. WCAG success criterion 3.1.4 (level AAA) asks for a way to find the expanded form, and most style guides ask writers to spell out an abbreviation the first time it is used.
A list of abbreviations at the start or end of a long report is the usual way to meet both needs. The HTML export produces one using the abbr element, which also shows the expansion as a tooltip in most browsers.
Limitations: what the result does not prove
- It recognises abbreviations with at least two capital letters. Lower-case abbreviations such as "e.g.", "approx." or "kg" are not listed.
- Definitions are found only in the two bracket patterns. A definition written as "COPD, which stands for..." or in a separate glossary is not recognised.
- The Schwartz and Hearst method matches letters, not meaning. It can miss a definition whose words do not contain the acronym's letters in order, and it can occasionally match the wrong words.
- Runs of four or more capitalised words are treated as headings or emphasis and skipped, and Roman numerals such as II and IV are left out unless you include them.
Privacy: where your data goes
Everything you paste, type or drop is processed in this browser tab. It is not uploaded, logged, stored or sent to analytics. Session recording and tag-manager scripts are switched off on this page.
Standards and sources
- WCAG 2.2 Understanding 3.1.4 Abbreviations - checked 19 Sep 2026
- Schwartz and Hearst (2003) - A simple algorithm for identifying abbreviation definitions in biomedical text
- HTML Living Standard - the abbr element
Frequently asked questions
What is the difference between an acronym and an abbreviation?
An abbreviation is any shortened form of a word or phrase. An acronym is an abbreviation made from initial letters and read as a word, such as NASA; an initialism is read letter by letter, such as BBC. This page lists all of them, as long as they contain at least two capital letters.
How do I find undefined acronyms in a long document?
Paste the text here and read the "Never defined" count and findings. Each undefined acronym is listed with the line of its first use and how many times it appears, so you can add the definition in the right place.
Should I define acronyms that everyone knows?
It depends on your audience and style guide. Many guides let writers skip very common ones such as UK or PDF. This page keeps a short list of widely known abbreviations and does not count them as undefined unless you switch that off.
Where should an acronym be spelled out?
At its first use in the main text, usually as the full words followed by the acronym in brackets. In long documents, many guides also repeat the definition at the first use in each chapter and provide a list of abbreviations.
How does the extractor match an acronym to its meaning?
It uses the Schwartz and Hearst algorithm: starting from the acronym's last letter, it looks backwards through the words next to the bracket for each letter in turn, requiring the first letter to start a word. The shortest run of words that contains all the letters in order is taken as the long form.
Can I export a list of abbreviations for my thesis or report?
Yes. Download CSV for a spreadsheet, Markdown for documentation, or HTML with abbr elements for a web page. Any expansion you type into the table is included, and the list is sorted alphabetically.
Last reviewed by the A2Z.Tools team against the sources listed above.
