fuzzyset

Search a set of strings/documents with fuzzy matching

import FuzzySet from 'fuzzyset';

const docs = ['afghanistan', 'albania', 'algeria', ...]

const set = new FuzzySet(docs);
const results = set.query('al');

for (const [i, score] of results) {
  console.log(countries[i], score);
}

// Albania 1.0
// Algeria 0.5
// ...

new FuzzySet(docs, options)

Instantiate a set, given docs and options.

docs

Array of strings, each of which could be a simple title, or the entire text of a document, or a list of keywords.

options

• synonyms - Object with keys as alias terms pointing to canonical terms. Ex: { nyc: 'New York City', philly: 'Philadelphia' }

• stopwords - Array of terms that should be ignored for having no semantic meaning. Ex: ['a', 'an', 'of', 'the', '&']

• tokenMatchFactor - Weight to give to matching normalized tokens. Default 0.5.

• tokenPrefixFactor - Weight to give to terms matching prefixes. Default 0.25.

• prefixFactor - Weight to give to direct document prefix term matching. Default 0.12

• vaccuumFactor - Weight to give to consonant terms with vowels removed. Default 0.8