Word Frequency Counter

How the text is broken into words

The text is lowercased, then every run of letters or digits in any script is taken as a word, with straight or curly apostrophes inside a word kept so contractions and possessives stay whole. Anything else, including hyphens, full stops and quotation marks, acts as a separator. The consequence is that hyphenated compounds are counted as their parts and decimal numbers as two numbers.

Because letters from every alphabet count, text in Cyrillic, Greek, Arabic or CJK scripts is counted word by word, and accented words in French, Spanish or German stay intact. Languages written without spaces between words, such as Chinese and Japanese, are counted as one word per run of characters, which is rarely what you want.

Worked example with the sample sentence

The sample text 'The quick brown fox jumps over the lazy dog. The dog was not amused.' has 14 total words and 11 unique words. The ranking starts with 'the' (3) and 'dog' (2); the other nine words each appear once and are listed alphabetically: amused, brown, fox, jumps, lazy, not, over, quick, was. The bar for 'dog' is two thirds the width of the bar for 'the', because bars are drawn relative to the top count.

Keyword density is count divided by total: 'the' makes up 3 of 14 words, or 21.4 percent, and 'dog' 14.3 percent.

How the list is sorted

• Highest count first.
• Ties are broken alphabetically, so among words that appear once 'amused' comes before 'brown'.
• There is no stop-word filter: 'the', 'and' and 'of' will usually top the list of any English text. Skip past them to the first content word.
• Long words are shown in a fixed-width column and may be cut off with an ellipsis in the display; the count beside them is still correct.

Reading the results for different jobs

Limits of this analysis

Counts are per single word, so a two-word phrase such as 'free tools' is not counted as a unit; look up each word separately and take the smaller figure as the upper bound. Plurals and verb forms are separate entries ('dog' and 'dogs' are two words), and there is no export button; copy the list by selecting it.