Repeated Phrase Finder

How Repeated Phrase Finder works

Repeated Phrase Finder handles case-insensitive repeated sequences between two and six words, ranking higher counts first and alphabetizing ties for deterministic output. The analyzer tokenizes Unicode words and numbers, slides an overlapping window of the selected width, counts identical normalized sequences in a map, and keeps counts above one. The Repeated Phrase Finder exposes its text, words per phrase controls so this particular convention stays inspectable instead of hiding behind an unexplained preset.

At width three, red blue green red blue green reports red blue green twice because that complete three-token window occurs at both ends. Editors can detect unintentional catchphrases, boilerplate, or repeated search language that a single-word frequency list cannot reveal. The Repeated Phrase Finder presents a deterministic plain-text result that can be copied and compared with the worked value before it enters another workflow.

Limits and private processing

Punctuation and capitalization are ignored, but word order is exact. Overlapping occurrences count separately, and semantic paraphrases are not considered matches. The Repeated Phrase Finder performs its complete operation locally in your browser; neither the source text nor its settings are uploaded, stored remotely, or sent through an external request.