Fuzzy Matching on One Column to then form Group IDs
I am trying to perform a fuzzy match of the “School Names” column in the attached workflow. There is no dictionary file, so the fuzzy match is on its own column. For values that are similar in name (say, +90% similar), then we would identify those as part of the same group for the purposes of a Group ID. The Group ID would increment for each group.
The workflow attached is inspired from two other similar discussions, but not quite taken to the finish line.
2 issues currently:
In my attached workflow, the Group ID isn’t incrementing correctly
The Jaro-Winkler component is not able to handle a dataset of 200K (ran for 2 hours, no progress)
The workflow aims to use fuzzy match, Postal Code, District, and State to determine what rows are part of the same group or not.
Appreciate any tips.