Search

Rank text by meaning

Each concept's top-ranked terms are matched as whole words in every chunk of the raw corpus, so nothing is tokenized or indexed ahead of time.

All searches →

How a search ranks text

A concept is its chosen term plus the terms that belong with it, each weighted on one scale: the chosen term is 100, and a term at 25 counts a quarter as much. Each chunk scores the sum of the weights of the concept terms it contains, so a chunk can rank high without the chosen term, through the terms that mean the same thing.

An example, with illustrative weights
  • xylophone
    100.0
  • glockenspiel
    26.2
  • marimba
    18.5
  • vibraphone
    12.1

Scoring three chunks with it:

100.0

The xylophone solo opened the show.

44.7

For the second movement the percussionist moved from the marimba to the glockenspiel.

—

The orchestra tuned up before the concert.

The second chunk never says “xylophone”, yet still ranks; the third shares no term with the concept and isn’t a hit. With several concepts, a chunk ranks high only if it fits every one.