How to use the word cloud generator
- Paste your text into the box, or choose a .txt file. The cloud and the table of counts appear as soon as there is text, and update as you type.
- Set Most words to show (5 to 200) and the shortest word to count. Leave common English words out, add your own words to leave out separated by commas, and choose whether numbers count as words and whether some words turn sideways.
- Press New arrangement to place the same words in a different layout. The size of each word never changes between arrangements, only its position.
- Use Download PNG for a picture, or Download SVG for a scalable drawing. The table underneath lists the 20 most common words and their counts, so the information does not depend on seeing the picture.
A short paragraph
Take the text "The dog and the fox. Dogs, DOGS and dogs were welcome. Fox-trot, fox trot, FOX." With common words removed and a shortest word length of 3, the word "dogs" counts 4, "fox" counts 4 and "trot" counts 2, because case is ignored and a hyphen splits "fox-trot" into two words. Words such as "the" and "and" are left out as common words.
In the longer sample text of the tool, "data" appears four times and several other words twice, so "data" is drawn largest. The size of a word grows with the square root of its share between the least and most common word shown, which keeps a few very common words from shrinking everything else to unreadable specks.
How words are counted
Text is split into words made of letters and digits, with an apostrophe kept inside a word, so don't and it's count as single words, including the curly apostrophe. Counting ignores case, so Dogs, DOGS and dogs are one word. Accented letters and non-Latin scripts count as letters.
The common word list holds about 150 plain English words such as the, of and was. It is a simple list, not a full dictionary, and it only applies to English. Turn it off for other languages, or add your own words to leave out. Word forms are not merged, so dog and dogs are counted separately.
How the cloud is laid out
Words are placed largest first on an outward spiral from the middle of the picture, and each one takes the first spot where it does not touch any word already placed. Because of this, the biggest words end up near the center. A word that finds no room is left out and named in the message, and a smaller count of words or a lower largest size would let more fit.
The layout was tested on 60 words and 250 words: no two words overlap, nothing leaves the picture, the same arrangement number always gives the same picture, and bigger counts always get bigger type. For a plain count without a picture, use the word counter, and to look at how hard a text is to read, the readability score calculator.
Privacy and accessibility
The text never leaves your browser. Words are escaped before they go into the picture, so markup in your text appears as plain words. The picture has a text description of its most common words, each word shows its count when you hover over it, and the table lists the counts in plain text. Colors come from a palette that stays distinct for common kinds of color blindness, but size and the table carry the meaning.
Assumptions and limits
- Counts are for the words as typed, with no stemming or merging of plural and singular forms.
- Only the first 2,000,000 characters of text are read. Larger files are refused.
- The common word list is English. Words that are not in it are never removed unless you add them.
- The picture is 900 by 560 units and is downloaded at 1.5 times that size as a PNG.
Frequently asked questions
Why is a word I expected missing from the cloud?
It was probably shorter than the shortest word setting, on the common word list, a number with numbers turned off, or one of the words you chose to leave out. A word that found no room is named in the message above the picture.
Can I make the same cloud again?
The same text and settings give the same counts and sizes. The arrangement depends on the arrangement number, which starts at the same value each time you open the page, so a fresh visit gives the same picture until you press New arrangement.
Is my text uploaded to make the cloud?
No. The words are counted and the picture is drawn on your device, and the downloads are created in your browser.
Does it work for languages other than English?
Yes for counting and drawing, since any letters and digits form words. The built-in list of common words is English only, so turn it off and list your own words to leave out.