Skip to content

Fuzzy Query

Returns documents that contain terms similar to the search term, as measured by a Levenshtein edit distance.

An edit distance is the number of one-character changes needed to turn one term into another. These changes can include:

  • Changing a character (e.g., box → fox)
  • Removing a character (e.g., black → lack)
  • Inserting a character (e.g., sic → sick)
  • Transposing two adjacent characters (e.g., act → cat)

To find similar terms, the fuzzy query creates a set of all possible variations, or expansions, of the search term within a specified edit distance. The query then returns exact matches for each expansion.

Example

import (
    es "github.com/elastic/go-elasticsearch/v8"
    "github.com/sdqri/effdsl/v2"
    eq "github.com/sdqri/effdsl/queries/existsquery"
)

query, err := effdsl.Define(
    effdsl.WithQuery(
        fq.FuzzyQuery(
            "user.id",
            "ki",
        ),
    ),
)

res, err := es.Search(
    es.Search.WithBody(strings.NewReader(query)),
)

Parameters

  • Field (string)
    (Required, positional) The field you wish to search. This is a required parameter.

  • Value (string)
    (Required, positional) The term you wish to find in the provided field. This is a required parameter.

  • WithFuzziness (string)
    (Optional, Functional option) The degree of fuzziness allowed for the search term (e.g., "AUTO", "1", "2", etc.). Defaults to no fuzziness.

  • WithMaxExpansions (int)
    (Optional, Functional option) Maximum number of terms to match. Defaults to 50.

  • WithPrefixLength (int)
    (Optional, Functional option) Number of initial characters that must match exactly. Defaults to 0.

  • WithTranspositions (bool)
    (Optional, Functional option) If true, allows transpositions of two adjacent characters. Defaults to true.

  • WithRewrite (Rewrite)
    (Optional, Functional option) Method used to rewrite the query. Valid values are:

    • constant_score: Query is rewritten to a constant score query.
    • scoring_boolean: Query is rewritten to a scoring boolean query.
    • constant_score_boolean: Query is rewritten to a constant score boolean query.
    • top_terms_N: Query is rewritten to match the top N scoring terms.
    • top_terms_boost_N: Query is rewritten to match the top N scoring terms with boosting.
    • top_terms_blended_freqs_N: Query is rewritten to match the top N scoring terms with blended frequencies.

Additional Information

For more details on the fuzzy query and its parameters, refer to the official Elasticsearch documentation on fuzzy queries.