BYTETOOLS

Keyword Extractor (RAKE)

Pull the key phrases out of any text with the RAKE algorithm. Ranks multi-word phrases by co-occurrence score, entirely in your browser.

This extractor is purely statistical. It runs the RAKE algorithm over the words you paste, using a built-in English stopword list — there is no AI, no network request, and nothing is sent anywhere.

1
4
1
20

What is the Keyword Extractor (RAKE)?

Keyword Extractor finds the phrases that define a piece of text using RAKE, the Rapid Automatic Keyword Extraction algorithm published by Rose and colleagues in 2010.

  • Full RAKE implementation with degree-over-frequency scoring
  • Finds multi-word key phrases, not just single words
  • Built-in English stopword list, roughly 180 words
  • Filter by phrase length and minimum occurrence count
  • Score bars make the ranking easy to scan
  • Export the ranked keywords as a CSV file

How to use the Keyword Extractor (RAKE)

  1. 1

    Paste the text you want to analyse into the input box.

  2. 2

    Set the shortest and longest phrase length you are interested in, in words.

  3. 3

    Raise the minimum occurrences if you only want phrases that repeat.

  4. 4

    Choose how many keywords to show, then read the ranked list.

  5. 5

    Copy the phrases or download the full table as CSV.

About the Keyword Extractor (RAKE)

Keyword Extractor finds the phrases that define a piece of text using RAKE, the Rapid Automatic Keyword Extraction algorithm published by Rose and colleagues in 2010. Candidate phrases are the runs of content words left when punctuation and common stopwords are stripped out, so it finds real multi-word terms rather than isolated words.

Each word is scored by its degree divided by its frequency in the phrase co-occurrence graph, and a phrase scores the sum of its words. Words that appear in many long phrases score higher, which is why a specific term like a three-word technical phrase outranks a common single word.

This is a statistical method with a built-in English stopword list — there is no AI and no network request, and the text you paste never leaves your browser. Tune the phrase length and minimum occurrences, then export the ranked list as CSV.

Frequently asked questions

What is RAKE keyword extraction?

RAKE stands for Rapid Automatic Keyword Extraction. It splits text into candidate phrases at punctuation and stopwords, then scores each word by how many other words it co-occurs with relative to how often it appears.

Why do longer phrases score higher?

A word's degree counts every word in every phrase it appears in, so words that keep turning up inside long phrases build a bigger score. That is deliberate — specific multi-word terms usually describe a document better than single words.

Is this the same as keyword density for SEO?

No. Keyword density just counts how often a term appears, whereas RAKE weighs how words combine into phrases. It is much better at surfacing the actual topics of a page than a raw frequency count.

Does it use AI or send my text anywhere?

Neither. It is a deterministic statistical algorithm running in your browser with a hand-written stopword list, so the same text always gives the same result and nothing is uploaded.

Why are some obvious keywords missing?

Anything on the stopword list, and any phrase shorter or longer than your length settings, is filtered out. Widen the phrase length range or drop the minimum occurrences to see more candidates.

Related tools