Find configurable hesitation and discourse filler phrases in text.
The frame below runs the same code as this page, in the reader's own browser. Nothing is sent to us, and nothing is sent to you.
Pick a dark background and the text and panels follow it, so the frame stays readable on a dark page.
A tokenized text is matched against a configurable list of discourse fillers and hesitation phrases, with counts and positions retained for review. Matching can be case-insensitive while still distinguishing a filler from a legitimate content word.
It was just really actually quite a good idea.
"just" - character 7 "really" - character 12 "actually" - character 19 "quite" - character 28
Transcript editing, presentation rehearsal, and language study find repeated fillers.