Fuzzy Address Matching

This node matches a set of string columns against a reference table and appends an identifying column (the ID column) of the closest match to the original table. The individual columns are compared with their counterparts in the reference table using the Levenshtein Distance and can be weighted.

Please note that at least 2 columns need to be selected.

Options

Comparison Columns

The columns to use for the fuzzy matching.

ID Column

The column from the reference table that identifies an address.

Column Weights

Comma-separated list of weights for the columns selected above. The higher the weight, the more important a column is for the selection of a match from the reference table.

Nodes

• Table Row to Variable5 ×
• Aggregated Distance2 ×
• Cell To Model2 ×
• Model to Cell2 ×
• Row Filter2 ×
• String Distances2 ×
• Ungroup2 ×
• Variable to Table Row2 ×
• Cell Splitter1 ×
• Column Filter Configuration1 ×
• Column Selection Configuration1 ×
• Component Input1 ×
• Component Output1 ×
• Extract Table Dimension1 ×
• GroupBy1 ×
• Math Formula1 ×
• Math Formula (Variable)1 ×
• Recursive Loop End1 ×
• Recursive Loop Start1 ×
• Similarity Search1 ×
• String Configuration1 ×
• String To Number1 ×

Extensions

• FeatureKNIME Base nodes
• FeatureKNIME Distance Matrix
• FeatureKNIME Ensemble Learning Wrappers
• FeatureKNIME Math Expression (JEP)
• FeatureKNIME Quick Forms

Links

• Fuzzy Lookup/Matching/Joining…by izaychik63 on 2023-10-04

  forum.knime.com/p/236744
• Identification of companies from questionnaires in a databaseby FVarlet on 2022-01-31

  forum.knime.com/p/146028