Help center Get started Knowledge center Product documentation API documentation
Product
Lens Weave Maskbits
Overview
Deidentification
Auto De-identification Excel De-identification Manual De-identification
Datasource
Datasource
Datasource Management
Data Model
Connectors Management
Admin
Admin
User Management Roles Management Group Management Organization Management
Product
Lens Weave Maskbits
Overview
Data
Data Import Output Result
Data Preparation
Cleansing Imputation Duplicates Filter Fuzzy Filter Outlier Treatment Binning Trimming Validation Sorting
Transformation
Data Transformation Bulk Data Transformation Merge Union Fuzzy Merge Encoding Normalization Scaling Standardization Aggregate Unpivot Functional Transformation SQL Executor
Statistical Plots
Box Plot Correlation Histogram Pair Wise Line Plot Scatter Plot Bar Plot Stem Plot Step Plot Fill Between Stack Plot Descriptive Statistics
Workflow Agent
Workflow Agent
Scheduler
Scheduler
Library
My Workflows Shared Workflows Audit
Datasource
Datasource
Datasource Management
Data Model
Connectors Management
Admin
Admin
User Management Roles Management Group Management Organization Management

Overview

The Fuzzy Filter node is a data preparation node in a Weave workflow that keeps rows whose column values approximately match the filter criteria, using a configurable fuzzy-matching algorithm. Where the standard Filter node compares values exactly, Fuzzy Filter accepts similarity-based matches and other approximate matches depending on the selected algorithm. The Filter Criteria popup uses the same Column / Filter Condition / Input / Action structure as the standard Filter; the difference is what “match” means inside the algorithm.

Fuzzy Filter nodes are added from the Data Preparation category of the node panel. Configuration is three-part: select the Filter Algorithm, configure the algorithm’s threshold value, and build filter criteria via the Fuzzy Filter Table + Filter button + Filter Criteria popup.

Note: “Weave” is the current name for what earlier interface versions called “Workflow.” This reference uses “Weave.”

When to use it

  • Customer records where the same entity appears with spelling variations (Smith / Smyth / Smithe).
  • Address or location data where city or region names vary (NYC / New York City / N.Y.C.).
  • Product catalogs where SKUs or names have typos or formatting differences.
  • Reconciling data from multiple sources where the same logical value is recorded inconsistently.
  • Cases where exact-match Filter misses semantically equivalent records.
  • Phonetic searches across name fields — Soundex Phonetic is specifically designed for English-language phonetic similarity.

Configuration

Filter Algorithm

Selects the algorithm used to compute similarity between values.

Soundex PhoneticPhonetic encoding — values encoding to the same Soundex code are considered matches. Designed for English-language phonetic similarity.English name matching, customer records with spelling variations, English place names.
Levenshtein distanceCounts the number of single-character edits (insertions, deletions, or replacements) needed to transform one string into another. Smaller edit counts indicate higher similarity.Typo detection, near-duplicate strings, names or values that differ by a few characters (e.g., “Smith” vs. “Smyth”, “Carretera” vs. “Carretra”).
Similarity ScoreCompares strings using word-based similarity that’s order-independent — strings with the same words in different orders score as similar.Multi-word values where word order varies (e.g., “Inter Quartile Range” vs. “Quartile Range Inter”), company or product names with reordered tokens, free-text fields with consistent vocabulary but inconsistent ordering.

Filter Criteria popup

Same UI as the standard Filter node — Column / Filter Condition / Input / Action structure. The key difference is that the Filter Condition operators apply fuzzy matching rather than exact comparison when the algorithm is configured.

ColumnUpstream column to filter on.
Filter ConditionSame 13-operator set as the standard Filter node (Equals, Not Equals, In, Not In, Less Than, Greater Than, etc.).
InputComparison value. Fuzzy matching applies the algorithm to compute similarity between the column’s value and the Input.
Action +Add another filter row.
ApplyCommits the criteria.

Key behaviors

Approximate matching, not exact. The fundamental difference between Fuzzy Filter and the standard Filter node is comparison semantics. Where the standard Filter requires exact equality (or strict ordering, or pattern), Fuzzy Filter measures similarity between the column’s value and the Input and admits rows above the algorithm’s threshold. This means “Smith” with Soundex Phonetic matches “Smyth” — both encode to the same Soundex code — where the standard Filter would treat them as distinct.

Algorithm determines what “similar” means. Different fuzzy algorithms measure similarity differently. Soundex Phonetic uses English phonetic encoding — good for names but poor for non-phonetic typos. The choice of algorithm is the most important configuration decision because it determines which kinds of variations will be matched.

Shared Filter Criteria UI. The Filter Criteria popup uses the same Column / Filter Condition / Input / Action structure as the standard Filter node and other places in the platform that surface a filter UI (Duplicates’ Filter Table, Lens Custom Script’s Filter operation). Designers familiar with the standard Filter UI use the same skills here.