PROCESIO
OverviewPlatform ActionsIntegrations & DemosCustom ActionsDeveloperโ€™s Guide

Untitled

This action searches for a given string inside a list of strings by estimating their similarity using configurable algorithms.&x20;

This action searches for a given string inside a list of strings by estimating their similarity using configurable algorithms.

Obs: Only the strings with a similarity score greater than or equal to the specified threshold are returned.

Inputs & Outputs

  • [ I ] List of strings to search into: Collection of strings in which the search will be performed.
  • [ I ] Similarity algorithm: Drop-down list of available string similarity algorithms.
    • Options include: BlockDistance, CosineSimilarity, JaccardSimilarity, DiceSimilarity, Jaro, JaroWinkler, Levenstein, MongeElkan, NeedlemanWunch, SmithWaterman, SmithWatermanGotoh, and others.
    • Default: SmithWaterman
  • [ I ] Similarity threshold [0-1]: Decimal number between 0 and 1. Strings with similarity below this value will not be included in the results.
    • Default: 0.85
  • [ I ] String to search for: The string that will be compared against the list items.
  • [ I ] Limit number of search results: Integer. Specifies how many top results to return.
    • Default: 100
    • If set to 0 โžœ returns all results.
  • [ O ] Output list of search results: List of strings from the input that are most similar to the searched string.

Example Usage

  • Drag Search by String Similarity onto the canvas.

  • Declare some variables to be used with the action:
    • myList โ†’ ["apple", "applesauce", "banana", "grape"]
    • mySearch โ†’ "apple"
    • myThreshold โ†’ 0.8
    • myAlgorithm โ†’ JaroWinkler

  • Assign the variables in the configuration:
    • List of strings to search into = myList
    • String to search for = mySearch
    • Similarity threshold = myThreshold
    • Similarity algorithm = myAlgorithm
    • Limit number of search results = 5
    • Output list of search results โ†’ similarStrings
  • Save & Run the process and then Check Instance to see the output.

Expected result: similarStrings = ["apple", "applesauce"]

Use Cases

  • Finding the closest matches for a search term inside a catalog or dataset.
  • Suggesting alternatives when user input is misspelled or partially correct.
  • Performing fuzzy search operations in process automations.

On this page